Pipeline run

0a257c2c-bc57-4d7c-a07c-41f4abf48f5e

Pipeline LLM cost (USD)

API 1: $0.0038 API 2: $0.0940 API 3: $0.0000 Total: $0.0978

Client output enrichment

v2 Skill cluster · Nature of work · AI index · Tech stack maturity · Evidence · KRA description

Nature of work · Video Codec Development

Build and optimize video codec components in C/C++ for x86 and ARM, including encoders/decoders, transcoding, motion estimation, rate control, and video pre/post-processing, with debugging, profiling, and standards work on AV1/H.265.

"Development and implementation of optimized algorithms for video encoders, video decoders, video pre and post processing components on x86 and ARM based CPUs"

Tech stack maturity

Mainstream Legacy cache hit

The stack centers on low-level C with SIMD optimization on ARM/NEON and SSE, which is typical of established performance-critical codec engineering rather than cloud-native or bleeding-edge AI tooling.

AI index (0 = no AI use, 5 = totally AI-dependent · v2.1)

0.50 / 5

· Title match

✓ Has AI skill

· AI skill (primary)

✓ AI skill (secondary)

· On AI team

· Builds AI products

vocab breakdown (legacy)

Assistants (×1): —

Frameworks (×2): —

Models / concepts (×3): Machine Learning

Evidence — skills matched in JD (19)

C C++ x86 SSE AVX ARM NEON Git AV1 H.265 H.264 MPEG-2 VP9 OpenCL CUDA Machine Learning Neural Networks Multithreading Cache Optimization

Skill cluster (5 dimension groups, role-scoped)

Codec Standards and Bitstreams

AV1 VP9

Hardware Acceleration and SIMD

SSE NEON

Video Codec Languages and DSLs

C C++

AI Governance and Model Security

Machine Learning

Cross-cutting / unaligned

x86 AVX ARM Git H.265 H.264 MPEG-2 OpenCL CUDA Neural Networks Multithreading Cache Optimization

Show KRA description ↓

The prospective candidate will be part of the Advanced Video and Research Team that designs and delivers video codec solutions for industry leaders in video technology. The key responsibilities of the job would be to deliver and excel on the following fronts (not limited to): • Development and implementation of optimized algorithms for video encoders, video decoders, video pre and post processing components on x86 and ARM based CPUs • Work involves implementation of high quality video encoders, decoders and transcoders and associated intellectual properties like Motion estimation, Rate Control algorithms, Scene Cut Detection, Fade-in / Fade-out Compensation, De-interlacing, De-noising as an example • Working on latest technology of Machine learning and Neural Network based video compression • Knowledge of C/C++ • Knowledge of x86 based development, intrinsic like SSE, AVX based coding • Knowledge of ARM based development, intrinsic like Neon coding • Debugging, profiling and development environments • Good knowledge of video standards like AV1 and H.265 • Working knowledge of H.264, MPEG-2 and VP9 is good to possess • Software Processes, Git, Configuration Management, Test Planning and Execution • Exposure to multi-threaded, cache optimal designs of video codecs • Exposure to OpenCL based GPU development / CUDA based programming • Aware of Machine learning and Neural Network basics.

Signals

Skill video-codec-engineer

0.26

Alias video-codec-engineer

1.00

KRA —

—

Post-classification

Centroid—

Alias collision log—

New-role queue—

New skills captured9

New KRA capturedyes

Captured for admin review

x86 primary ↔ Video Codec Engineer pending

AVX primary ↔ Video Codec Engineer pending

H.265 primary ↔ Video Codec Engineer pending

H.264 ↔ Video Codec Engineer pending

MPEG-2 ↔ Video Codec Engineer pending

OpenCL ↔ Video Codec Engineer pending

CUDA ↔ Video Codec Engineer pending

Neural Networks ↔ Video Codec Engineer pending

Cache Optimization ↔ Video Codec Engineer pending

R&R fragment (sim 0.00) ↔ Video Codec Engineer pending

Status: extract_details_done Created: 2026-05-20T14:32:55.161633Z Updated: 2026-05-20T14:37:47.431468Z

Flow Current 3-step pipeline

1 POST /skills/extract-from-jd

2 POST /skills/extract-details

3 POST /skills/final-role-output

Role Chosen role & resolution

Video Codec Engineer

CASE A

slug: video-codec-engineer · id: 22 · source: db

The primary skills emphasize video codec languages and hardware acceleration, fitting the role of a Video Codec Engineer.

Job description

Role : Video Codec Engineer 

Required Experience: Candidates must have development experience ranging from 2 to 4 years.
• Experience in implementing video compression standards based and/or proprietary Image and Video codecs/algorithms
• Must have exposure and development experience ARM and/or x86 based platforms like Xeon E5/E3, Core-i7/i5
• Experience of development using operating systems like Windows / Linux / OS X

Job Description: The prospective candidate will be part of the Advanced Video and Research Team that designs and delivers video codec solutions for industry leaders in video technology.

Responsibility: The key responsibilities of the job would be to deliver and excel on the following fronts (not limited to):
• Development and implementation of optimized algorithms for video encoders, video decoders, video pre and post processing components on x86 and ARM based CPUs
• Work involves implementation of high quality video encoders, decoders and transcoders and associated intellectual properties like Motion estimation, Rate Control algorithms, Scene Cut Detection, Fade-in / Fade-out Compensation, De-interlacing, De-noising as an example
• Working on latest technology of Machine learning and Neural Network based video compression

Educational Qualification: Masters or Bachelor’s Degree in Computer Science / Electronics and Communication

Required Technical Skills:
• Knowledge of C/C++
• Knowledge of x86 based development, intrinsic like SSE, AVX based coding
• Knowledge of ARM based development, intrinsic like Neon coding
• Debugging, profiling and development environments
• Good knowledge of video standards like AV1 and H.265
• Working knowledge of H.264, MPEG-2 and VP9 is good to possess
• Software Processes, Git, Configuration Management, Test Planning and Execution
• Exposure to multi-threaded, cache optimal designs of video codecs
• Exposure to OpenCL based GPU development / CUDA based programming
• Aware of Machine learning and Neural Network basics.

Location: Bengaluru, Karnataka

Skills from this JD

Each row merges API 1 extraction, API 2 library match / v3 orchestration (dimensions + locked dims), and API 3 persistence tags.

C Primary Library skill Existing skill (matched library)

Canonical: C id=4 · c

Aliases — catalog

C# (CANONICAL) primary
C (CANONICAL)
C# 1 (VERSION)
C# 10 (VERSION)
C# 11 (VERSION)
C# 12 (VERSION)
C# 13 (VERSION)
C# 14 (VERSION)
C# 2 (VERSION)
C# 3 (VERSION)
C# 4 (VERSION)
C# 5 (VERSION)
C# 6 (VERSION)
C# 7 (VERSION)
C# 8 (VERSION)
C# 9 (VERSION)
C# latest (VERSION)
C#1 (VERSION)
C#10 (VERSION)
C#11 (VERSION)
C#12 (VERSION)
C#2 (VERSION)
C#3 (VERSION)
C#4 (VERSION)
C#5 (VERSION)
C#6 (VERSION)
C#7 (VERSION)
C#8 (VERSION)
C#9 (VERSION)
C++ (CANONICAL)
C++03 (VERSION)
C++11 (VERSION)
C++14 (VERSION)
C++17 (VERSION)
C++20 (VERSION)
C++23 (VERSION)
C++26 (VERSION)
C++98 (VERSION)
c sharp (VERSION)
c# (VERSION)
cpp03 (VERSION)
cpp11 (VERSION)
cpp14 (VERSION)
cpp17 (VERSION)
cpp20 (VERSION)
cpp23 (VERSION)
cpp26 (VERSION)
cpp98 (VERSION)
csharp (VERSION)
modern C++ (VERSION)

Context tags (catalog)

.NET ASP.NET Azure Boost C# 9 C11 C99 CLR CMake Entity Framework LINQ MSBuild MVC Makefile NuGet Qt RAII Roslyn STL TDD Unity Visual Studio WPF Web API WinForms Xamarin async/await attributes buffer overflow constexpr delegates dependency injection dynamic memory events free function pointer gcc gdb header files inline interfaces linker malloc memory leak memory management move semantics multithreading operator overloading pointer pointers preprocessor segmentation fault smart pointers stdio.h stdlib.h struct syntax error templates typedef xUnit

Stored enrichment (catalog DB)

Category: Language
Sub-category: Programming Language
Vendor: Microsoft
License: mit
Year introduced: 2000
Confidence: 0.99
Version strategy: SEPARATE_ENTITY
Version tag: latest

Maturity reasoning: C# is a mainstream hiring staple with high JD volume across .NET, Azure, and enterprise roles; Microsoft continues active platform investment in .NET, reinforcing broad adoption.

Skill profile (library / DB)

Skill nature: LANGUAGE
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 6
Sub-category id: 96
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Cross-Platform App Languages Catalog dimension db id 167

Library dimension (catalog)

Roles linked in library: Hybrid Mobile Developer
Programming Languages Catalog dimension db id 1

Library dimension (catalog)

Roles linked in library: Backend Engineer, Full Stack Engineer
Programming Languages for ML Systems Catalog dimension db id 39

Library dimension (catalog)

Roles linked in library: ML Engineer, ML Ops Engineer
Programming Languages for XR Catalog dimension db id 97

Library dimension (catalog)

Roles linked in library: AR/VR Engineer
React Frontend Development Catalog dimension db id 96

Library dimension (catalog)
Video Codec Languages and DSLs Catalog dimension db id 225

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

C++ Primary Library skill Existing skill (matched library)

Canonical: C id=4 · c

Aliases — catalog

C# (CANONICAL) primary
C (CANONICAL)
C# 1 (VERSION)
C# 10 (VERSION)
C# 11 (VERSION)
C# 12 (VERSION)
C# 13 (VERSION)
C# 14 (VERSION)
C# 2 (VERSION)
C# 3 (VERSION)
C# 4 (VERSION)
C# 5 (VERSION)
C# 6 (VERSION)
C# 7 (VERSION)
C# 8 (VERSION)
C# 9 (VERSION)
C# latest (VERSION)
C#1 (VERSION)
C#10 (VERSION)
C#11 (VERSION)
C#12 (VERSION)
C#2 (VERSION)
C#3 (VERSION)
C#4 (VERSION)
C#5 (VERSION)
C#6 (VERSION)
C#7 (VERSION)
C#8 (VERSION)
C#9 (VERSION)
C++ (CANONICAL)
C++03 (VERSION)
C++11 (VERSION)
C++14 (VERSION)
C++17 (VERSION)
C++20 (VERSION)
C++23 (VERSION)
C++26 (VERSION)
C++98 (VERSION)
c sharp (VERSION)
c# (VERSION)
cpp03 (VERSION)
cpp11 (VERSION)
cpp14 (VERSION)
cpp17 (VERSION)
cpp20 (VERSION)
cpp23 (VERSION)
cpp26 (VERSION)
cpp98 (VERSION)
csharp (VERSION)
modern C++ (VERSION)

Context tags (catalog)

Stored enrichment (catalog DB)

Category: Language
Sub-category: Programming Language
Vendor: Microsoft
License: mit
Year introduced: 2000
Confidence: 0.99
Version strategy: SEPARATE_ENTITY
Version tag: latest

Maturity reasoning: C# is a mainstream hiring staple with high JD volume across .NET, Azure, and enterprise roles; Microsoft continues active platform investment in .NET, reinforcing broad adoption.

Skill profile (library / DB)

Skill nature: LANGUAGE
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 6
Sub-category id: 96
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Cross-Platform App Languages Catalog dimension db id 167

Library dimension (catalog)

Roles linked in library: Hybrid Mobile Developer
Programming Languages Catalog dimension db id 1

Library dimension (catalog)

Roles linked in library: Backend Engineer, Full Stack Engineer
Programming Languages for ML Systems Catalog dimension db id 39

Library dimension (catalog)

Roles linked in library: ML Engineer, ML Ops Engineer
Programming Languages for XR Catalog dimension db id 97

Library dimension (catalog)

Roles linked in library: AR/VR Engineer
React Frontend Development Catalog dimension db id 96

Library dimension (catalog)
Video Codec Languages and DSLs Catalog dimension db id 225

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

x86 Primary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity well_known confidence 0.95

x86 remains a dominant ISA in server, desktop, and systems roles; job postings for low-level, OS, and performance engineering commonly mention x86/x86-64, and Intel/AMD roadmaps continue active support.

Vendor & license

(0.90)

Context keywords

x64 IA-32 assembly language microarchitecture instruction set registers pipeline cache coherence virtual memory system calls memory management multithreading performance optimization debugging hardware compatibility

Ambiguity flagged

Could be confused with: arm, arm64

“x86” in JDs can be confused with other CPU instruction set architectures like ARM/ARM64, especially when describing target architectures.

Versioning

Not versioned

Type assignment

Architecture ·instruction_set_architecture confidence 0.97

x86 is fundamentally an instruction set architecture, and the Architecture vs Concept rule applies because it describes a system shape for how software is built and executed rather than a knowledge unit.

Derived legacy fields

Category: Architecture
Sub-category: instruction_set_architecture
Skill nature: PATTERN
Volatility: STABLE
Typical lifespan: EVERGREEN
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

Locked dimensions (v3 placement)

x86 Instruction Set Architecture
Pipeline tentative id

Covers the x86 CPU instruction set, execution modes, and low-level architectural behavior used when writing or optimizing code for Intel/AMD processors. This skill belongs here because x86 is the core ISA target for codec assembly, SIMD use, and platform-specific performance work.

SSE Primary Library skill Existing skill (matched library)

Canonical: SSE id=1718 · sse

Aliases — catalog

SSE (CANONICAL) primary

Context tags (catalog)

CPU architecture SIMD assembly language cache optimization compiler optimizations data-level parallelism hardware acceleration intrinsics low-level programming memory alignment multi-threading parallel processing performance optimization register allocation vectorization

Stored enrichment (catalog DB)

Category: Concept
Sub-category: Simd Instruction Set Extension
Confidence: 0.98
Version strategy: NOT_APPLICABLE

Maturity reasoning: SSE is a long-established x86 SIMD extension; it appears in systems/performance JDs and is widely supported by compilers and CPUs, though newer AVX/AVX2/AVX-512 often supersede it for greenfield optimization.

Skill profile (library / DB)

Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 2
Sub-category id: 1277
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Hardware Acceleration and SIMD Catalog dimension db id 235

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

AVX Primary Library skill Existing skill (matched library)

Canonical: AVX2 id=1716 · avx2

Aliases — catalog

AVX2 (CANONICAL) primary

Context tags (catalog)

AVX-512 FMA SIMD bit manipulation cache utilization compiler optimizations data-level parallelism floating-point operations instruction pipelining intrinsics microarchitecture multithreading parallel processing performance optimization vectorization

Stored enrichment (catalog DB)

Category: Concept
Sub-category: Simd Instruction Set Extension
Confidence: 0.98
Version strategy: NOT_APPLICABLE

Maturity reasoning: AVX2 appears in specialized systems/performance JDs, but far less often than mainstream platforms; it’s a CPU SIMD extension used in HPC, media, and low-level optimization rather than a broad hiring staple.

Skill profile (library / DB)

Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 2
Sub-category id: 1277
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Hardware Acceleration and SIMD Catalog dimension db id 235

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

ARM Primary Library skill Existing skill (matched library)

Canonical: ARM id=1621 · arm

Aliases — catalog

AArch64 (VERSION)
ARM (CANONICAL)
ARMv8 (VERSION)
arm64 (VERSION)

Context tags (catalog)

ARM Cortex ARMv7 ARMv8 JTAG NEON RISC SoC Thumb assembly language cross-compilation debugging embedded systems hardware abstraction low-level programming microcontroller real-time operating systems

Stored enrichment (catalog DB)

Category: Architecture
Sub-category: Instruction Set Architecture
Confidence: 0.91
Version strategy: SEPARATE_ENTITY
Version tag: ARMv8-A

Maturity reasoning: ARM is a dominant instruction-set architecture in mobile, embedded, and increasingly server/cloud chips; job postings commonly mention ARM64/AArch64 alongside Linux and systems work.

Skill profile (library / DB)

Skill nature: PATTERN
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 1
Sub-category id: 1222
Extractable: True
Also category: False

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

NEON Primary Library skill Existing skill (matched library)

Canonical: NEON id=1717 · neon

Aliases — catalog

NEON (CANONICAL) primary

Context tags (catalog)

SIMD algorithm optimization bit manipulation compiler support data parallelism embedded systems floating-point operations hardware acceleration instruction set low-level programming memory alignment multithreading parallel processing performance optimization vectorization

Stored enrichment (catalog DB)

Category: Concept
Sub-category: Simd Instruction Set Extension
Confidence: 0.98
Version strategy: NOT_APPLICABLE

Maturity reasoning: ARM NEON is a standard SIMD extension on mobile/embedded ARM chips and appears in many performance/embedded JDs and compiler docs, especially for multimedia and ML acceleration.

Skill profile (library / DB)

Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 2
Sub-category id: 1277
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Hardware Acceleration and SIMD Catalog dimension db id 235

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

Git Primary Library skill Existing skill (matched library)

Canonical: Git id=1002 · git

Aliases — catalog

Git (CANONICAL)

Context tags (catalog)

CI/CD GitHub GitLab branching checkout clone commit fork merging pull request rebase remote repository stash versioning

Stored enrichment (catalog DB)

Category: Tool
Sub-category: Version Control Tool
Vendor: Linus Torvalds
License: gpl_v2
Year introduced: 2005
Confidence: 0.99
Version strategy: NOT_APPLICABLE

Maturity reasoning: Git is a hiring-pipeline staple: it appears in the vast majority of software engineering job descriptions and is the default VCS on GitHub/GitLab/Bitbucket.

Skill profile (library / DB)

Skill nature: TOOL
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 13
Sub-category id: 730
Extractable: True
Also category: False

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

AV1 Primary Library skill Existing skill (matched library)

Canonical: AV1 id=1678 · av1

Aliases — catalog

AV1 (CANONICAL) primary

Context tags (catalog)

AV1 decoder AV1 encoder HEVC adaptive streaming bitrate codec container formats decoding encoding interoperability media playback streaming video compression video delivery video quality

Stored enrichment (catalog DB)

Category: Standard
Sub-category: Video Codec Standard
Vendor: Alliance for Open Media
License: other_open
Year introduced: 2018
Confidence: 0.96
Version strategy: NOT_APPLICABLE

Maturity reasoning: AV1 is increasingly requested in streaming/media JDs and supported by major vendors (YouTube, Netflix, Chrome/Firefox), but it’s still far less universal than H.264/H.265 in job postings.

Skill profile (library / DB)

Skill nature: STANDARD
Volatility: EMERGING
Typical lifespan: EVERGREEN
Category id: 12
Sub-category id: 1308
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Codec Standards and Bitstreams Catalog dimension db id 227

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

H.265 Primary Library skill Existing skill (matched library)

Canonical: H.265/HEVC id=1677 · h-265-hevc

Aliases — catalog

H.265/HEVC (CANONICAL) primary

Context tags (catalog)

4K HDR bitrate codec container format decoding encoding hardware acceleration inter-frame intra-frame level profile streaming video compression video quality

Stored enrichment (catalog DB)

Category: Standard
Sub-category: Video Codec Standard
Vendor: ITU-T
License: unknown
Year introduced: 2013
Confidence: 0.97
Version strategy: NOT_APPLICABLE

Maturity reasoning: Widely used in streaming, broadcast, and device pipelines; job ads for video/codec engineers still mention HEVC alongside H.264/AV1, and major vendors ship hardware decode/encode support.

Skill profile (library / DB)

Skill nature: STANDARD
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 12
Sub-category id: 1308
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Codec Standards and Bitstreams Catalog dimension db id 227

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

H.264 Secondary Library skill Existing skill (matched library)

Canonical: H.264/AVC id=1676 · h-264-avc

Aliases — catalog

H.264/AVC (CANONICAL) primary

Context tags (catalog)

AVC B-frame CABAC HLS I-frame MP4 P-frame bitrate container format entropy coding level profile streaming video compression video encoding

Stored enrichment (catalog DB)

Category: Standard
Sub-category: Video Codec Standard
Vendor: ITU-T
License: unknown
Year introduced: 2003
Confidence: 0.97
Version strategy: NOT_APPLICABLE

Maturity reasoning: H.264/AVC is still widely required in video streaming, conferencing, and hardware encoding/decoding JDs; it remains a default codec in major vendor stacks despite newer alternatives like AV1.

Skill profile (library / DB)

Skill nature: STANDARD
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 12
Sub-category id: 1308
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Codec Standards and Bitstreams Catalog dimension db id 227

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

MPEG-2 Secondary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity deprecated confidence 0.93

MPEG-2 is largely superseded in new deployments by H.264/H.265 and AV1; recent job postings rarely list it except in legacy broadcast/video systems, and modern vendor docs focus on newer codecs.

Vendor & license

(0.95)

Context keywords

video compression bitstream MPEG-2 Transport Stream MPEG-2 Program Stream interlaced video frame rate encoding decoding streaming digital television multimedia video quality error resilience bitrate container format

Ambiguity low

MPEG-2 is a specific video compression standard; typical JDs won’t confuse it with other unrelated skills.

Versioning

Not versioned

Type assignment

Concept ·general confidence 0.00

Stage 4 failed; fallback typed record.

Derived legacy fields

Category: Concept
Sub-category: general
Skill nature: CONCEPT
Volatility: DEPRECATED
Typical lifespan: SHORT_LIVED
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

Locked dimensions (v3 placement)

Video Compression Standards
Pipeline tentative id

Standards and specifications for encoding and decoding digital video streams. MPEG-2 belongs here because it defines a widely used video compression format, bitstream syntax, and interoperability rules for broadcast and disc media.

VP9 Secondary Library skill Existing skill (matched library)

Canonical: VP9 id=1679 · vp9

Aliases — catalog

VP9 (CANONICAL) primary

Context tags (catalog)

AV1 WebM adaptive streaming bitrate codec optimization decoding encoding hardware acceleration lossless compression media container real-time streaming video compression video conferencing video quality video streaming

Stored enrichment (catalog DB)

Category: Standard
Sub-category: Video Codec Standard
Vendor: Google
License: bsd
Year introduced: 2013
Confidence: 0.95
Version strategy: NOT_APPLICABLE

Maturity reasoning: VP9 appears in some media/streaming and browser JDs, but far less often than H.264/AV1; market demand is limited and it’s largely overshadowed by AV1 in new deployments.

Skill profile (library / DB)

Skill nature: STANDARD
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 12
Sub-category id: 1308
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Codec Standards and Bitstreams Catalog dimension db id 227

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

OpenCL Secondary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity niche confidence 0.86

OpenCL still appears in GPU/HPC job postings, but far less often than CUDA or vendor SDKs; GitHub activity is steady yet modest, indicating specialized use rather than broad hiring demand.

Vendor & license

Khronos Group ·other_open ·since 2008 (0.95)

Context keywords

GPU parallelism kernels cl_device_id cl_mem cl_command_queue cl_program cl_context compute units OpenCL C platforms device drivers memory management task parallelism heterogeneous computing

Ambiguity low

OpenCL is a specific parallel computing API/language; typical JDs won’t confuse it with other common skills in the catalog.

Versioning

Not versioned

Type assignment

Language ·parallel_computing_api_language confidence 0.90

OpenCL is best treated as a programming language/API specification for expressing parallel kernels and host code, so it fits the Language type rather than a tool or framework.

Derived legacy fields

Category: Language
Sub-category: parallel_computing_api_language
Skill nature: LANGUAGE
Volatility: STABLE
Typical lifespan: EVERGREEN
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)
Accelerators and Hardware for ML Catalog dimension db id 58

Library dimension (catalog)

Roles linked in library: ML Engineer
Accelerators and Hardware for ML Catalog dimension db id 58

Library dimension (catalog)

Roles linked in library: ML Engineer

Locked dimensions (v3 placement)

GPU Compute Programming
Pipeline tentative id

Programming models and APIs for writing parallel compute kernels that run on GPUs and other accelerators. OpenCL belongs here because it defines portable kernel execution, memory management, and host-device coordination for heterogeneous compute.
Accelerators and Hardware for ML
Reuses catalog slug

Specialized hardware and accelerator programming used to offload compute-intensive workloads. OpenCL can fit here when used as a portable GPU/accelerator interface, though the dimension is broader than ML and includes hardware-aware execution.
Accelerators and Hardware for ML
Reuses catalog slug

Specialized hardware and accelerators for training and serving machine learning models.

CUDA Secondary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity well_known confidence 0.93

CUDA appears in many ML/HPC job descriptions and is the de facto NVIDIA GPU programming stack; NVIDIA continues active platform support and ecosystem investment, indicating broad market adoption.

Vendor & license

NVIDIA ·proprietary ·since 2006 (0.95)

Context keywords

NVIDIA GPU parallel computing CUDA cores cuDNN TensorRT CUDA Toolkit kernel memory management CUDA streams thrust OpenCL GPGPU performance optimization compute capability

Ambiguity low

CUDA is a specific NVIDIA GPU programming platform; typical JDs won’t confuse it with other distinct skills in the catalog.

Versioning

Not versioned

Type assignment

Language ·gpu_programming_language confidence 0.90

CUDA is fundamentally a programming language/toolchain for expressing GPU kernels and device code, so it fits the Language category rather than a library or framework.

Derived legacy fields

Category: Language
Sub-category: gpu_programming_language
Skill nature: LANGUAGE
Volatility: STABLE
Typical lifespan: EVERGREEN
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

Accelerators and Hardware for ML Catalog dimension db id 58

Library dimension (catalog)

Roles linked in library: ML Engineer

Locked dimensions (v3 placement)

GPU Accelerators and Parallel Compute
Reuses catalog slug

Programming and optimization work that targets GPUs and other accelerators for high-throughput compute. CUDA belongs here because it is the primary programming model for NVIDIA GPU execution, memory movement, and kernel optimization.

Machine Learning Secondary Library skill Existing skill (matched library)

Canonical: Machine Learning id=1356 · machine-learning

Aliases — catalog

Machine Learning (CANONICAL)

Context tags (catalog)

Keras PyTorch TensorFlow cross-validation data preprocessing ensemble methods feature engineering hyperparameter tuning model evaluation natural language processing neural networks reinforcement learning scikit-learn supervised learning unsupervised learning

Stored enrichment (catalog DB)

Category: Concept
Sub-category: Machine Learning
Confidence: 0.98
Version strategy: NOT_APPLICABLE

Maturity reasoning: Machine Learning appears in large volumes of job descriptions across data, product, and platform roles, and major cloud vendors (AWS, Google Cloud, Azure) offer dedicated ML services and certifications, indicating broad adoption.

Skill profile (library / DB)

Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 2
Sub-category id: 1024
Extractable: True
Also category: False

Dimensions (API 2 worklist)

AI Governance and Model Security Catalog dimension db id 50

Library dimension (catalog)

Roles linked in library: AI Engineer, ML Engineer, ML Ops Engineer
React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

Neural Networks Secondary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity well_known confidence 0.98

Broadly listed in ML/AI job descriptions and core to modern deep learning stacks; major frameworks like PyTorch and TensorFlow center on neural networks.

Vendor & license

(0.95)

Context keywords

backpropagation activation functions convolutional layers recurrent networks dropout gradient descent TensorFlow Keras PyTorch overfitting training data hyperparameters transfer learning model architecture regularization

Ambiguity low

“Neural Networks” is a specific ML model concept; typical JDs won’t confuse it with other distinct skills in the catalog.

Versioning

Not versioned

Type assignment

Concept ·machine_learning_model_concept confidence 0.97

Neural Networks are a named knowledge unit in machine learning, so by the Concept vs Methodology rule they are a Concept rather than an Architecture or Framework.

Derived legacy fields

Category: Concept
Sub-category: machine_learning_model_concept
Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

React Frontend Development Catalog dimension db id 96

Library dimension (catalog)

Locked dimensions (v3 placement)

Neural Network Fundamentals
Pipeline tentative id

Core concepts, architectures, and training behavior of artificial neural networks. This fits the target skill because it refers to the model family itself rather than a specific deployment, hardware, or operations stack.

Multithreading Secondary Library skill Existing skill (matched library)

Canonical: multithreading id=82 · multithreading

Aliases — catalog

multithreading (CANONICAL) primary

Context tags (catalog)

atomic operations concurrent execution condition variable deadlock lock contention mutex non-blocking parallelism producer-consumer race condition semaphore synchronization thread pool thread-safe work stealing

Stored enrichment (catalog DB)

Category: Concept
Sub-category: Concurrency Concept
Confidence: 0.94
Version strategy: NOT_APPLICABLE

Maturity reasoning: Common requirement in JDs for backend, systems, and mobile roles; widely taught and used across Java, C++, Go, and Python concurrency stacks.

Skill profile (library / DB)

Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Category id: 2
Sub-category id: 7
Extractable: True
Also category: False

Dimensions (API 2 worklist)

Concurrency and Parallel Processing Catalog dimension db id 17

Library dimension (catalog)

Roles linked in library: Backend Engineer

Cache Optimization Secondary New / orchestrated New / unmatched skill (orchestrated in API 2)

Skill enrichment (orchestrator / LLM)

Maturity well_known confidence 0.84

Common performance topic in JDs for backend, systems, and mobile roles; cache-miss reduction and CPU cache locality are standard interview and profiling concerns across major stacks.

Vendor & license

(0.95)

Context keywords

cache eviction cache hit ratio cache coherence LRU memcached Redis CDN data locality prefetching write-through write-back cache partitioning distributed caching cache warming performance tuning

Ambiguity low

“Cache Optimization” is a specific performance-tuning concept; it’s unlikely to be confused with other distinct catalog skills in typical JDs.

Versioning

Not versioned

Type assignment

Concept ·cache_optimization confidence 0.93

This is best treated as a Concept because it names a technical knowledge unit about improving cache behavior, not a specific tool, methodology, or architecture.

Derived legacy fields

Category: Concept
Sub-category: cache_optimization
Skill nature: CONCEPT
Volatility: STABLE
Typical lifespan: EVERGREEN
Version strategy: NOT_APPLICABLE

Dimensions (API 2 worklist)

Codec Performance Benchmarking Catalog dimension db id 238

Library dimension (catalog)

Roles linked in library: Video Codec Engineer
React Frontend Development Catalog dimension db id 96

Library dimension (catalog)
Codec Performance Benchmarking Catalog dimension db id 238

Library dimension (catalog)

Roles linked in library: Video Codec Engineer

Locked dimensions (v3 placement)

Codec Performance Benchmarking
Reuses catalog slug

Measurement and profiling practices used to improve codec speed, memory use, and throughput. Cache optimization belongs here because it is a core performance-tuning technique for reducing stalls and improving encode/decode efficiency.
Memory Locality Optimization
Pipeline tentative id

Techniques for improving how software uses CPU caches and memory hierarchy. Cache Optimization fits here because it specifically targets locality, reduced misses, and better access patterns in performance-critical systems.
Codec Performance Benchmarking
Reuses catalog slug

Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.

Library artifacts (this run)

No artifact rows for this run.

nano JD Parser — gpt-4.1-nano click to toggle

RoleVideo Codec Engineer

ExperienceCandidates must have development experience ranging from 2 to 4 years.

DomainOther

Location Bengaluru, India

JD type pass

Show raw JSON

{
  "JD_type": "pass",
  "about_company": null,
  "certifications": [],
  "company_name": null,
  "ctc": null,
  "domain": {
    "primary": {
      "aliases": [],
      "domain": "Other"
    },
    "secondary": null
  },
  "education": [
    {
      "level": "Bachelor\u0027s",
      "qualification": "BTECH/BE/MTECH/ME - Computer Science / Electronics and Communication",
      "raw": "Masters or Bachelor\u2019s Degree in Computer Science / Electronics and Communication",
      "requirement": "required"
    }
  ],
  "experience": {
    "max": 4,
    "min": 2,
    "raw": "Candidates must have development experience ranging from 2 to 4 years."
  },
  "job_locations": [
    {
      "aliases": [
        "Bangalore"
      ],
      "city": "Bengaluru",
      "country": "India",
      "state": "Karnataka",
      "work_mode": null
    }
  ],
  "role": "Video Codec Engineer",
  "role_aliases": [
    "Codec Engineer",
    "Video Engineer",
    "Video Compression Engineer"
  ],
  "role_archetype": "Engineering",
  "roles_and_responsibilities": [
    {
      "bullet_count": 0,
      "heading": "Job Description",
      "heading_was_present": true,
      "source_marker": {
        "first_5_words": "The prospective candidate will be",
        "last_5_words": "in video technology."
      },
      "text": "The prospective candidate will be part of the Advanced Video and Research Team that designs and delivers video codec solutions for industry leaders in video technology.",
      "word_count": 25
    },
    {
      "bullet_count": 3,
      "heading": "Responsibility",
      "heading_was_present": true,
      "source_marker": {
        "first_5_words": "The key responsibilities of the",
        "last_5_words": "video compression"
      },
      "text": "The key responsibilities of the job would be to deliver and excel on the following fronts (not limited to):\n\u2022 Development and implementation of optimized algorithms for video encoders, video decoders, video pre and post processing components on x86 and ARM based CPUs\n\u2022 Work involves implementation of high quality video encoders, decoders and transcoders and associated intellectual properties like Motion estimation, Rate Control algorithms, Scene Cut Detection, Fade-in / Fade-out Compensation, De-interlacing, De-noising as an example\n\u2022 Working on latest technology of Machine learning and Neural Network based video compression",
      "word_count": 64
    },
    {
      "bullet_count": 10,
      "heading": "Required Technical Skills",
      "heading_was_present": true,
      "source_marker": {
        "first_5_words": "\u2022 Knowledge of C/C++\n\u2022 Knowledge",
        "last_5_words": "and Neural Network basics."
      },
      "text": "\u2022 Knowledge of C/C++\n\u2022 Knowledge of x86 based development, intrinsic like SSE, AVX based coding\n\u2022 Knowledge of ARM based development, intrinsic like Neon coding\n\u2022 Debugging, profiling and development environments\n\u2022 Good knowledge of video standards like AV1 and H.265\n\u2022 Working knowledge of H.264, MPEG-2 and VP9 is good to possess\n\u2022 Software Processes, Git, Configuration Management, Test Planning and Execution\n\u2022 Exposure to multi-threaded, cache optimal designs of video codecs\n\u2022 Exposure to OpenCL based GPU development / CUDA based programming\n\u2022 Aware of Machine learning and Neural Network basics.",
      "word_count": 104
    }
  ],
  "urls": []
}

API 1 — extract-from-jd click to toggle

{
  "final_skills": [
    {
      "is_primary": true,
      "skill_name": "C"
    },
    {
      "is_primary": true,
      "skill_name": "C++"
    },
    {
      "is_primary": true,
      "skill_name": "x86"
    },
    {
      "is_primary": true,
      "skill_name": "SSE"
    },
    {
      "is_primary": true,
      "skill_name": "AVX"
    },
    {
      "is_primary": true,
      "skill_name": "ARM"
    },
    {
      "is_primary": true,
      "skill_name": "NEON"
    },
    {
      "is_primary": true,
      "skill_name": "Git"
    },
    {
      "is_primary": true,
      "skill_name": "AV1"
    },
    {
      "is_primary": true,
      "skill_name": "H.265"
    },
    {
      "is_primary": false,
      "skill_name": "H.264"
    },
    {
      "is_primary": false,
      "skill_name": "MPEG-2"
    },
    {
      "is_primary": false,
      "skill_name": "VP9"
    },
    {
      "is_primary": false,
      "skill_name": "OpenCL"
    },
    {
      "is_primary": false,
      "skill_name": "CUDA"
    },
    {
      "is_primary": false,
      "skill_name": "Machine Learning"
    },
    {
      "is_primary": false,
      "skill_name": "Neural Networks"
    },
    {
      "is_primary": false,
      "skill_name": "Multithreading"
    },
    {
      "is_primary": false,
      "skill_name": "Cache Optimization"
    }
  ],
  "jd_role": {
    "display_name": "Video Codec Engineer",
    "rationale": null,
    "role_aliases": [
      "Codec Engineer",
      "Video Engineer",
      "Video Compression Engineer"
    ],
    "role_archetype": "Engineering",
    "slug": ""
  },
  "nano_parsed": {
    "JD_type": "pass",
    "about_company": null,
    "certifications": [],
    "company_name": null,
    "ctc": null,
    "domain": {
      "primary": {
        "aliases": [],
        "domain": "Other"
      },
      "secondary": null
    },
    "education": [
      {
        "level": "Bachelor\u0027s",
        "qualification": "BTECH/BE/MTECH/ME - Computer Science / Electronics and Communication",
        "raw": "Masters or Bachelor\u2019s Degree in Computer Science / Electronics and Communication",
        "requirement": "required"
      }
    ],
    "experience": {
      "max": 4,
      "min": 2,
      "raw": "Candidates must have development experience ranging from 2 to 4 years."
    },
    "job_locations": [
      {
        "aliases": [
          "Bangalore"
        ],
        "city": "Bengaluru",
        "country": "India",
        "state": "Karnataka",
        "work_mode": null
      }
    ],
    "role": "Video Codec Engineer",
    "role_aliases": [
      "Codec Engineer",
      "Video Engineer",
      "Video Compression Engineer"
    ],
    "role_archetype": "Engineering",
    "roles_and_responsibilities": [
      {
        "bullet_count": 0,
        "heading": "Job Description",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "The prospective candidate will be",
          "last_5_words": "in video technology."
        },
        "text": "The prospective candidate will be part of the Advanced Video and Research Team that designs and delivers video codec solutions for industry leaders in video technology.",
        "word_count": 25
      },
      {
        "bullet_count": 3,
        "heading": "Responsibility",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "The key responsibilities of the",
          "last_5_words": "video compression"
        },
        "text": "The key responsibilities of the job would be to deliver and excel on the following fronts (not limited to):\n\u2022 Development and implementation of optimized algorithms for video encoders, video decoders, video pre and post processing components on x86 and ARM based CPUs\n\u2022 Work involves implementation of high quality video encoders, decoders and transcoders and associated intellectual properties like Motion estimation, Rate Control algorithms, Scene Cut Detection, Fade-in / Fade-out Compensation, De-interlacing, De-noising as an example\n\u2022 Working on latest technology of Machine learning and Neural Network based video compression",
        "word_count": 64
      },
      {
        "bullet_count": 10,
        "heading": "Required Technical Skills",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "\u2022 Knowledge of C/C++\n\u2022 Knowledge",
          "last_5_words": "and Neural Network basics."
        },
        "text": "\u2022 Knowledge of C/C++\n\u2022 Knowledge of x86 based development, intrinsic like SSE, AVX based coding\n\u2022 Knowledge of ARM based development, intrinsic like Neon coding\n\u2022 Debugging, profiling and development environments\n\u2022 Good knowledge of video standards like AV1 and H.265\n\u2022 Working knowledge of H.264, MPEG-2 and VP9 is good to possess\n\u2022 Software Processes, Git, Configuration Management, Test Planning and Execution\n\u2022 Exposure to multi-threaded, cache optimal designs of video codecs\n\u2022 Exposure to OpenCL based GPU development / CUDA based programming\n\u2022 Aware of Machine learning and Neural Network basics.",
        "word_count": 104
      }
    ],
    "urls": []
  },
  "rejected": false,
  "rejection_reason": null,
  "run_id": "0a257c2c-bc57-4d7c-a07c-41f4abf48f5e",
  "stage3_signals": {
    "alias_found": true,
    "alias_match_roles": [
      {
        "display_name": "Video Codec Engineer",
        "matched_count": null,
        "role_id": 22,
        "score": 1.0,
        "slug": "video-codec-engineer",
        "total_count": null
      }
    ],
    "kra_match_roles": [],
    "skill_match_roles": [
      {
        "display_name": "Video Codec Engineer",
        "matched_count": 5,
        "role_id": 22,
        "score": 0.2632,
        "slug": "video-codec-engineer",
        "total_count": 19
      },
      {
        "display_name": "ML Engineer",
        "matched_count": 2,
        "role_id": 3,
        "score": 0.1053,
        "slug": "ml-engineer",
        "total_count": 19
      },
      {
        "display_name": "Backend Engineer",
        "matched_count": 2,
        "role_id": 1,
        "score": 0.1053,
        "slug": "backend-engineer",
        "total_count": 19
      },
      {
        "display_name": "ML Ops Engineer",
        "matched_count": 2,
        "role_id": 16,
        "score": 0.1053,
        "slug": "ml-ops-engineer",
        "total_count": 19
      },
      {
        "display_name": "AI Engineer",
        "matched_count": 1,
        "role_id": 13,
        "score": 0.0526,
        "slug": "ai-engineer",
        "total_count": 19
      }
    ]
  },
  "stage4_decision": {
    "alias_collision_detected": false,
    "case": "A",
    "chosen_role": {
      "display_name": "Video Codec Engineer",
      "matched_count": null,
      "role_id": 22,
      "score": 1.0,
      "slug": "video-codec-engineer",
      "total_count": null
    },
    "confidence": 0.8,
    "is_new_role": false,
    "llm2_fired": false,
    "llm2_reasoning": null,
    "new_role_display_name": null,
    "new_role_slug": null,
    "queued": false,
    "reasoning": "Stage 1 title \u0027Video Codec Engineer\u0027 (name match, sim 1.00); KRA inconclusive"
  },
  "stage5_updates": {
    "centroid_n_after": null,
    "centroid_updated": false,
    "collision_log_id": null,
    "new_kra_attached": {
      "best_kra_similarity": 0.0,
      "queue_id": 57,
      "r_and_r_preview": "The prospective candidate will be part of the Advanced Video and Research Team that designs and delivers video codec solutions for industry leaders in video technology.\n\nThe key responsibilities of th",
      "role_display_name": "Video Codec Engineer",
      "role_slug": "video-codec-engineer",
      "status": "pending"
    },
    "new_skills_attached": [
      {
        "is_primary": true,
        "queue_id": 1584,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "x86",
        "status": "pending"
      },
      {
        "is_primary": true,
        "queue_id": 1585,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "AVX",
        "status": "pending"
      },
      {
        "is_primary": true,
        "queue_id": 1586,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "H.265",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1587,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "H.264",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1588,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "MPEG-2",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1589,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "OpenCL",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1590,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "CUDA",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1591,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "Neural Networks",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 1592,
        "role_display_name": "Video Codec Engineer",
        "role_slug": "video-codec-engineer",
        "skill_name": "Cache Optimization",
        "status": "pending"
      }
    ],
    "queue_entry_id": null,
    "v3_pipeline_triggered": false,
    "v3_role_slug": null,
    "v3_run_id": null
  }
}

API 2 — extract-details

{
  "alias_matches": [
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1609,
      "existing_alias_text": "C",
      "input_term": "C",
      "matched_canonical": {
        "category_id": 6,
        "display_name": "C",
        "id": 4,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "c",
        "sub_category_id": 96,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 418,
      "existing_alias_text": "C++",
      "input_term": "C++",
      "matched_canonical": {
        "category_id": 6,
        "display_name": "C",
        "id": 4,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "c",
        "sub_category_id": 96,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2688,
      "existing_alias_text": "SSE",
      "input_term": "SSE",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "SSE",
        "id": 1718,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "sse",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "TODO: REMOVE AFTER TESTING \u2014 alias DB write disabled",
      "alias_persisted": false,
      "existing_alias_id": 2686,
      "existing_alias_text": "AVX2",
      "input_term": "AVX",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "AVX2",
        "id": 1716,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "avx2",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "embedding_alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2573,
      "existing_alias_text": "ARM",
      "input_term": "ARM",
      "matched_canonical": {
        "category_id": 1,
        "display_name": "ARM",
        "id": 1621,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PATTERN",
        "slug": "arm",
        "sub_category_id": 1222,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2687,
      "existing_alias_text": "NEON",
      "input_term": "NEON",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "NEON",
        "id": 1717,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "neon",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1613,
      "existing_alias_text": "Git",
      "input_term": "Git",
      "matched_canonical": {
        "category_id": 13,
        "display_name": "Git",
        "id": 1002,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "git",
        "sub_category_id": 730,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2648,
      "existing_alias_text": "AV1",
      "input_term": "AV1",
      "matched_canonical": {
        "category_id": 12,
        "display_name": "AV1",
        "id": 1678,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "av1",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "TODO: REMOVE AFTER TESTING \u2014 alias DB write disabled",
      "alias_persisted": false,
      "existing_alias_id": 2647,
      "existing_alias_text": "H.265/HEVC",
      "input_term": "H.265",
      "matched_canonical": {
        "category_id": 12,
        "display_name": "H.265/HEVC",
        "id": 1677,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "h-265-hevc",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "embedding_alias"
    },
    {
      "alias_persist_skipped_reason": "TODO: REMOVE AFTER TESTING \u2014 alias DB write disabled",
      "alias_persisted": false,
      "existing_alias_id": 2646,
      "existing_alias_text": "H.264/AVC",
      "input_term": "H.264",
      "matched_canonical": {
        "category_id": 12,
        "display_name": "H.264/AVC",
        "id": 1676,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "h-264-avc",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "embedding_alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2649,
      "existing_alias_text": "VP9",
      "input_term": "VP9",
      "matched_canonical": {
        "category_id": 12,
        "display_name": "VP9",
        "id": 1679,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "vp9",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2015,
      "existing_alias_text": "Machine Learning",
      "input_term": "Machine Learning",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "Machine Learning",
        "id": 1356,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "machine-learning",
        "sub_category_id": 1024,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 223,
      "existing_alias_text": "multithreading",
      "input_term": "Multithreading",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "multithreading",
        "id": 82,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "multithreading",
        "sub_category_id": 7,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    }
  ],
  "candidate_roles": [
    {
      "display_name": "Hybrid Mobile Developer",
      "id": 11,
      "rationale": null,
      "role_archetype": null,
      "slug": "hybrid-mobile-developer",
      "source": "db"
    },
    {
      "display_name": "Backend Engineer",
      "id": 1,
      "rationale": null,
      "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
      "slug": "backend-engineer",
      "source": "db"
    },
    {
      "display_name": "Full Stack Engineer",
      "id": 15,
      "rationale": null,
      "role_archetype": null,
      "slug": "full-stack-engineer",
      "source": "db"
    },
    {
      "display_name": "ML Engineer",
      "id": 3,
      "rationale": null,
      "role_archetype": null,
      "slug": "ml-engineer",
      "source": "db"
    },
    {
      "display_name": "ML Ops Engineer",
      "id": 16,
      "rationale": null,
      "role_archetype": null,
      "slug": "ml-ops-engineer",
      "source": "db"
    },
    {
      "display_name": "AR/VR Engineer",
      "id": 8,
      "rationale": null,
      "role_archetype": null,
      "slug": "ar-vr-engineer",
      "source": "db"
    },
    {
      "display_name": "Video Codec Engineer",
      "id": 22,
      "rationale": null,
      "role_archetype": null,
      "slug": "video-codec-engineer",
      "source": "db"
    },
    {
      "display_name": "AI Engineer",
      "id": 13,
      "rationale": null,
      "role_archetype": null,
      "slug": "ai-engineer",
      "source": "db"
    }
  ],
  "chosen_role": {
    "display_name": "Video Codec Engineer",
    "id": 22,
    "rationale": "The primary skills emphasize video codec languages and hardware acceleration, fitting the role of a Video Codec Engineer.",
    "role_archetype": null,
    "slug": "video-codec-engineer",
    "source": "db"
  },
  "dimensions": [
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Cross-Platform App Languages",
        "id": 167,
        "rationale": "Languages used to implement shared mobile features across iOS and Android from a common codebase. This is the primary coding surface for hybrid app logic, UI behavior, and platform-specific branching.",
        "slug": "cross-platform-app-languages",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Hybrid Mobile Developer",
          "id": 11,
          "rationale": null,
          "role_archetype": null,
          "slug": "hybrid-mobile-developer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages",
        "id": 1,
        "rationale": "Primary implementation languages used to build client and server feature code. Full stack engineers need enough fluency to move across layers and implement product behavior end to end.",
        "slug": "programming-languages",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Backend Engineer",
          "id": 1,
          "rationale": null,
          "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
          "slug": "backend-engineer",
          "source": "db"
        },
        {
          "display_name": "Full Stack Engineer",
          "id": 15,
          "rationale": null,
          "role_archetype": null,
          "slug": "full-stack-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages for ML Systems",
        "id": 39,
        "rationale": "Languages used to build training code, inference services, evaluation jobs, and ML glue code. This is the primary implementation surface for ML engineers across experimentation and productionization.",
        "slug": "programming-languages-for-ml-systems",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        },
        {
          "display_name": "ML Ops Engineer",
          "id": 16,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-ops-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages for XR",
        "id": 97,
        "rationale": "Primary implementation languages used to build immersive client features, interaction logic, and device-specific runtime behavior. This is the core coding surface for AR/VR experiences.",
        "slug": "programming-languages-for-xr",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AR/VR Engineer",
          "id": 8,
          "rationale": null,
          "role_archetype": null,
          "slug": "ar-vr-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Video Codec Languages and DSLs",
        "id": 225,
        "rationale": "Implementation languages and codec-specific description languages used to build compression, decoding, and transform logic. This is the primary coding surface for codec behavior and bitstream manipulation.",
        "slug": "video-codec-languages-and-dsls",
        "source": "db"
      },
      "input_skill": "C",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Cross-Platform App Languages",
        "id": 167,
        "rationale": "Languages used to implement shared mobile features across iOS and Android from a common codebase. This is the primary coding surface for hybrid app logic, UI behavior, and platform-specific branching.",
        "slug": "cross-platform-app-languages",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Hybrid Mobile Developer",
          "id": 11,
          "rationale": null,
          "role_archetype": null,
          "slug": "hybrid-mobile-developer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages",
        "id": 1,
        "rationale": "Primary implementation languages used to build client and server feature code. Full stack engineers need enough fluency to move across layers and implement product behavior end to end.",
        "slug": "programming-languages",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Backend Engineer",
          "id": 1,
          "rationale": null,
          "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
          "slug": "backend-engineer",
          "source": "db"
        },
        {
          "display_name": "Full Stack Engineer",
          "id": 15,
          "rationale": null,
          "role_archetype": null,
          "slug": "full-stack-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages for ML Systems",
        "id": 39,
        "rationale": "Languages used to build training code, inference services, evaluation jobs, and ML glue code. This is the primary implementation surface for ML engineers across experimentation and productionization.",
        "slug": "programming-languages-for-ml-systems",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        },
        {
          "display_name": "ML Ops Engineer",
          "id": 16,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-ops-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages for XR",
        "id": 97,
        "rationale": "Primary implementation languages used to build immersive client features, interaction logic, and device-specific runtime behavior. This is the core coding surface for AR/VR experiences.",
        "slug": "programming-languages-for-xr",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AR/VR Engineer",
          "id": 8,
          "rationale": null,
          "role_archetype": null,
          "slug": "ar-vr-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Video Codec Languages and DSLs",
        "id": 225,
        "rationale": "Implementation languages and codec-specific description languages used to build compression, decoding, and transform logic. This is the primary coding surface for codec behavior and bitstream manipulation.",
        "slug": "video-codec-languages-and-dsls",
        "source": "db"
      },
      "input_skill": "C++",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Hardware Acceleration and SIMD",
        "id": 235,
        "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
        "slug": "hardware-acceleration-and-simd",
        "source": "db"
      },
      "input_skill": "SSE",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Hardware Acceleration and SIMD",
        "id": 235,
        "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
        "slug": "hardware-acceleration-and-simd",
        "source": "db"
      },
      "input_skill": "AVX",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "ARM",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Hardware Acceleration and SIMD",
        "id": 235,
        "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
        "slug": "hardware-acceleration-and-simd",
        "source": "db"
      },
      "input_skill": "NEON",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "Git",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Standards and Bitstreams",
        "id": 227,
        "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
        "slug": "codec-standards-and-bitstreams",
        "source": "db"
      },
      "input_skill": "AV1",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Standards and Bitstreams",
        "id": 227,
        "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
        "slug": "codec-standards-and-bitstreams",
        "source": "db"
      },
      "input_skill": "H.265",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Standards and Bitstreams",
        "id": 227,
        "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
        "slug": "codec-standards-and-bitstreams",
        "source": "db"
      },
      "input_skill": "H.264",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Standards and Bitstreams",
        "id": 227,
        "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
        "slug": "codec-standards-and-bitstreams",
        "source": "db"
      },
      "input_skill": "VP9",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "AI Governance and Model Security",
        "id": 50,
        "rationale": "Controls and documentation used to make models safer, auditable, and compliant. ML engineers use this to manage model risk, supply chain integrity, and governance requirements.",
        "slug": "ai-governance-and-model-security",
        "source": "db"
      },
      "input_skill": "Machine Learning",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        },
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        },
        {
          "display_name": "ML Ops Engineer",
          "id": 16,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-ops-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "Machine Learning",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Concurrency and Parallel Processing",
        "id": 17,
        "rationale": "Programming techniques for handling multiple requests and background work safely and efficiently. Includes synchronization, async execution, and coordination of concurrent tasks.",
        "slug": "concurrency-and-parallel-processing",
        "source": "db"
      },
      "input_skill": "Multithreading",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Backend Engineer",
          "id": 1,
          "rationale": null,
          "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
          "slug": "backend-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "x86",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "MPEG-2",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "OpenCL",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Accelerators and Hardware for ML",
        "id": 58,
        "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
        "slug": "accelerators-and-hardware-for-ml",
        "source": "db"
      },
      "input_skill": "OpenCL",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Accelerators and Hardware for ML",
        "id": 58,
        "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
        "slug": "accelerators-and-hardware-for-ml",
        "source": "db"
      },
      "input_skill": "OpenCL",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Accelerators and Hardware for ML",
        "id": 58,
        "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
        "slug": "accelerators-and-hardware-for-ml",
        "source": "db"
      },
      "input_skill": "CUDA",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "ML Engineer",
          "id": 3,
          "rationale": null,
          "role_archetype": null,
          "slug": "ml-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "Neural Networks",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Performance Benchmarking",
        "id": 238,
        "rationale": "Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.",
        "slug": "codec-performance-benchmarking",
        "source": "db"
      },
      "input_skill": "Cache Optimization",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "React Frontend Development",
        "id": 96,
        "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
        "slug": "d_init_01",
        "source": "db"
      },
      "input_skill": "Cache Optimization",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Codec Performance Benchmarking",
        "id": 238,
        "rationale": "Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.",
        "slug": "codec-performance-benchmarking",
        "source": "db"
      },
      "input_skill": "Cache Optimization",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Video Codec Engineer",
          "id": 22,
          "rationale": null,
          "role_archetype": null,
          "slug": "video-codec-engineer",
          "source": "db"
        }
      ]
    }
  ],
  "input_final_skills": [
    "C",
    "C++",
    "x86",
    "SSE",
    "AVX",
    "ARM",
    "NEON",
    "Git",
    "AV1",
    "H.265",
    "H.264",
    "MPEG-2",
    "VP9",
    "OpenCL",
    "CUDA",
    "Machine Learning",
    "Neural Networks",
    "Multithreading",
    "Cache Optimization"
  ],
  "input_llm_skills": [
    "C",
    "C++",
    "x86",
    "SSE",
    "AVX",
    "ARM",
    "NEON",
    "Git",
    "AV1",
    "H.265",
    "H.264",
    "MPEG-2",
    "VP9",
    "OpenCL",
    "CUDA",
    "Machine Learning",
    "Neural Networks",
    "Multithreading",
    "Cache Optimization"
  ],
  "new_aliases_persisted": 0,
  "run_id": "0a257c2c-bc57-4d7c-a07c-41f4abf48f5e",
  "skills_detail": [
    {
      "aliases_in_db": [
        {
          "alias_text": "C",
          "alias_type": "CANONICAL",
          "id": 1609,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#",
          "alias_type": "CANONICAL",
          "id": 52,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 1",
          "alias_type": "VERSION",
          "id": 53,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 10",
          "alias_type": "VERSION",
          "id": 62,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 11",
          "alias_type": "VERSION",
          "id": 63,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 12",
          "alias_type": "VERSION",
          "id": 64,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 13",
          "alias_type": "VERSION",
          "id": 65,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 14",
          "alias_type": "VERSION",
          "id": 66,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 2",
          "alias_type": "VERSION",
          "id": 54,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 3",
          "alias_type": "VERSION",
          "id": 55,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 4",
          "alias_type": "VERSION",
          "id": 56,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 5",
          "alias_type": "VERSION",
          "id": 57,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 6",
          "alias_type": "VERSION",
          "id": 58,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 7",
          "alias_type": "VERSION",
          "id": 59,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 8",
          "alias_type": "VERSION",
          "id": 60,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 9",
          "alias_type": "VERSION",
          "id": 61,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# latest",
          "alias_type": "VERSION",
          "id": 1655,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#1",
          "alias_type": "VERSION",
          "id": 2228,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#10",
          "alias_type": "VERSION",
          "id": 2246,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#11",
          "alias_type": "VERSION",
          "id": 2247,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#12",
          "alias_type": "VERSION",
          "id": 2248,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#2",
          "alias_type": "VERSION",
          "id": 2229,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#3",
          "alias_type": "VERSION",
          "id": 2230,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#4",
          "alias_type": "VERSION",
          "id": 2231,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#5",
          "alias_type": "VERSION",
          "id": 2232,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#6",
          "alias_type": "VERSION",
          "id": 2233,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#7",
          "alias_type": "VERSION",
          "id": 2234,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#8",
          "alias_type": "VERSION",
          "id": 2235,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#9",
          "alias_type": "VERSION",
          "id": 2236,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++",
          "alias_type": "CANONICAL",
          "id": 418,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++03",
          "alias_type": "VERSION",
          "id": 420,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++11",
          "alias_type": "VERSION",
          "id": 421,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++14",
          "alias_type": "VERSION",
          "id": 422,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++17",
          "alias_type": "VERSION",
          "id": 423,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++20",
          "alias_type": "VERSION",
          "id": 424,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++23",
          "alias_type": "VERSION",
          "id": 425,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++26",
          "alias_type": "VERSION",
          "id": 426,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++98",
          "alias_type": "VERSION",
          "id": 427,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp03",
          "alias_type": "VERSION",
          "id": 879,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp11",
          "alias_type": "VERSION",
          "id": 880,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp14",
          "alias_type": "VERSION",
          "id": 881,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp17",
          "alias_type": "VERSION",
          "id": 882,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp20",
          "alias_type": "VERSION",
          "id": 883,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp23",
          "alias_type": "VERSION",
          "id": 884,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp26",
          "alias_type": "VERSION",
          "id": 885,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp98",
          "alias_type": "VERSION",
          "id": 886,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "modern C++",
          "alias_type": "VERSION",
          "id": 428,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 6,
        "display_name": "C",
        "id": 4,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "c",
        "sub_category_id": 96,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Cross-Platform App Languages",
            "id": 167,
            "rationale": "Languages used to implement shared mobile features across iOS and Android from a common codebase. This is the primary coding surface for hybrid app logic, UI behavior, and platform-specific branching.",
            "slug": "cross-platform-app-languages",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Hybrid Mobile Developer",
              "id": 11,
              "rationale": null,
              "role_archetype": null,
              "slug": "hybrid-mobile-developer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages",
            "id": 1,
            "rationale": "Primary implementation languages used to build client and server feature code. Full stack engineers need enough fluency to move across layers and implement product behavior end to end.",
            "slug": "programming-languages",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Backend Engineer",
              "id": 1,
              "rationale": null,
              "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
              "slug": "backend-engineer",
              "source": "db"
            },
            {
              "display_name": "Full Stack Engineer",
              "id": 15,
              "rationale": null,
              "role_archetype": null,
              "slug": "full-stack-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages for ML Systems",
            "id": 39,
            "rationale": "Languages used to build training code, inference services, evaluation jobs, and ML glue code. This is the primary implementation surface for ML engineers across experimentation and productionization.",
            "slug": "programming-languages-for-ml-systems",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            },
            {
              "display_name": "ML Ops Engineer",
              "id": 16,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-ops-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages for XR",
            "id": 97,
            "rationale": "Primary implementation languages used to build immersive client features, interaction logic, and device-specific runtime behavior. This is the core coding surface for AR/VR experiences.",
            "slug": "programming-languages-for-xr",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AR/VR Engineer",
              "id": 8,
              "rationale": null,
              "role_archetype": null,
              "slug": "ar-vr-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": []
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Video Codec Languages and DSLs",
            "id": 225,
            "rationale": "Implementation languages and codec-specific description languages used to build compression, decoding, and transform logic. This is the primary coding surface for codec behavior and bitstream manipulation.",
            "slug": "video-codec-languages-and-dsls",
            "source": "db"
          },
          "input_skill": "C",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "C",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "C",
          "alias_type": "CANONICAL",
          "id": 1609,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#",
          "alias_type": "CANONICAL",
          "id": 52,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 1",
          "alias_type": "VERSION",
          "id": 53,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 10",
          "alias_type": "VERSION",
          "id": 62,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 11",
          "alias_type": "VERSION",
          "id": 63,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 12",
          "alias_type": "VERSION",
          "id": 64,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 13",
          "alias_type": "VERSION",
          "id": 65,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 14",
          "alias_type": "VERSION",
          "id": 66,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 2",
          "alias_type": "VERSION",
          "id": 54,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 3",
          "alias_type": "VERSION",
          "id": 55,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 4",
          "alias_type": "VERSION",
          "id": 56,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 5",
          "alias_type": "VERSION",
          "id": 57,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 6",
          "alias_type": "VERSION",
          "id": 58,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 7",
          "alias_type": "VERSION",
          "id": 59,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 8",
          "alias_type": "VERSION",
          "id": 60,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# 9",
          "alias_type": "VERSION",
          "id": 61,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C# latest",
          "alias_type": "VERSION",
          "id": 1655,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#1",
          "alias_type": "VERSION",
          "id": 2228,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#10",
          "alias_type": "VERSION",
          "id": 2246,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#11",
          "alias_type": "VERSION",
          "id": 2247,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#12",
          "alias_type": "VERSION",
          "id": 2248,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#2",
          "alias_type": "VERSION",
          "id": 2229,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#3",
          "alias_type": "VERSION",
          "id": 2230,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#4",
          "alias_type": "VERSION",
          "id": 2231,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#5",
          "alias_type": "VERSION",
          "id": 2232,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#6",
          "alias_type": "VERSION",
          "id": 2233,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#7",
          "alias_type": "VERSION",
          "id": 2234,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#8",
          "alias_type": "VERSION",
          "id": 2235,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C#9",
          "alias_type": "VERSION",
          "id": 2236,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++",
          "alias_type": "CANONICAL",
          "id": 418,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++03",
          "alias_type": "VERSION",
          "id": 420,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++11",
          "alias_type": "VERSION",
          "id": 421,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++14",
          "alias_type": "VERSION",
          "id": 422,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++17",
          "alias_type": "VERSION",
          "id": 423,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++20",
          "alias_type": "VERSION",
          "id": 424,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++23",
          "alias_type": "VERSION",
          "id": 425,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++26",
          "alias_type": "VERSION",
          "id": 426,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "C++98",
          "alias_type": "VERSION",
          "id": 427,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp03",
          "alias_type": "VERSION",
          "id": 879,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp11",
          "alias_type": "VERSION",
          "id": 880,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp14",
          "alias_type": "VERSION",
          "id": 881,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp17",
          "alias_type": "VERSION",
          "id": 882,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp20",
          "alias_type": "VERSION",
          "id": 883,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp23",
          "alias_type": "VERSION",
          "id": 884,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp26",
          "alias_type": "VERSION",
          "id": 885,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "cpp98",
          "alias_type": "VERSION",
          "id": 886,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "modern C++",
          "alias_type": "VERSION",
          "id": 428,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 6,
        "display_name": "C",
        "id": 4,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "c",
        "sub_category_id": 96,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Cross-Platform App Languages",
            "id": 167,
            "rationale": "Languages used to implement shared mobile features across iOS and Android from a common codebase. This is the primary coding surface for hybrid app logic, UI behavior, and platform-specific branching.",
            "slug": "cross-platform-app-languages",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Hybrid Mobile Developer",
              "id": 11,
              "rationale": null,
              "role_archetype": null,
              "slug": "hybrid-mobile-developer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages",
            "id": 1,
            "rationale": "Primary implementation languages used to build client and server feature code. Full stack engineers need enough fluency to move across layers and implement product behavior end to end.",
            "slug": "programming-languages",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Backend Engineer",
              "id": 1,
              "rationale": null,
              "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
              "slug": "backend-engineer",
              "source": "db"
            },
            {
              "display_name": "Full Stack Engineer",
              "id": 15,
              "rationale": null,
              "role_archetype": null,
              "slug": "full-stack-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages for ML Systems",
            "id": 39,
            "rationale": "Languages used to build training code, inference services, evaluation jobs, and ML glue code. This is the primary implementation surface for ML engineers across experimentation and productionization.",
            "slug": "programming-languages-for-ml-systems",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            },
            {
              "display_name": "ML Ops Engineer",
              "id": 16,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-ops-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages for XR",
            "id": 97,
            "rationale": "Primary implementation languages used to build immersive client features, interaction logic, and device-specific runtime behavior. This is the core coding surface for AR/VR experiences.",
            "slug": "programming-languages-for-xr",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AR/VR Engineer",
              "id": 8,
              "rationale": null,
              "role_archetype": null,
              "slug": "ar-vr-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": []
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Video Codec Languages and DSLs",
            "id": 225,
            "rationale": "Implementation languages and codec-specific description languages used to build compression, decoding, and transform logic. This is the primary coding surface for codec behavior and bitstream manipulation.",
            "slug": "video-codec-languages-and-dsls",
            "source": "db"
          },
          "input_skill": "C++",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "C++",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "x86",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "x86",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Architecture",
          "skill_nature": "PATTERN",
          "sub_category": "instruction_set_architecture",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "STABLE"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": true,
            "confused_with": [
              "arm",
              "arm64"
            ],
            "reasoning": "\u201cx86\u201d in JDs can be confused with other CPU instruction set architectures like ARM/ARM64, especially when describing target architectures."
          },
          "context_keywords": {
            "context_keywords": [
              "x64",
              "IA-32",
              "assembly language",
              "microarchitecture",
              "instruction set",
              "registers",
              "pipeline",
              "cache coherence",
              "virtual memory",
              "system calls",
              "memory management",
              "multithreading",
              "performance optimization",
              "debugging",
              "hardware compatibility"
            ]
          },
          "maturity": {
            "confidence": 0.95,
            "maturity": "well_known",
            "reasoning": "x86 remains a dominant ISA in server, desktop, and systems roles; job postings for low-level, OS, and performance engineering commonly mention x86/x86-64, and Intel/AMD roadmaps continue active support."
          },
          "skill_id": "x86",
          "vendor_license": {
            "confidence": 0.9,
            "license": null,
            "vendor": null,
            "year_introduced": null
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [],
        "locked_dimensions": [
          {
            "description": "Covers the x86 CPU instruction set, execution modes, and low-level architectural behavior used when writing or optimizing code for Intel/AMD processors. This skill belongs here because x86 is the core ISA target for codec assembly, SIMD use, and platform-specific performance work.",
            "exemplar_skills": [
              "x86",
              "x86-64",
              "IA-32",
              "SSE",
              "AVX",
              "AVX2",
              "AVX-512",
              "inline assembly",
              "CPU feature detection"
            ],
            "in_scope": "x86, x86-64, IA-32, instruction encoding, registers, addressing modes, calling conventions, SIMD instruction sets, SSE, AVX, AVX2, AVX-512, inline assembly, CPU feature detection",
            "name": "x86 Instruction Set Architecture",
            "out_of_scope": "ARM and AArch64 targets, GPU kernels, compiler construction, codec benchmarking, defect analysis, and general systems programming not tied to the x86 ISA",
            "overlap_flags": [
              {
                "reason": "x86 is often used to optimize codec hot paths, but benchmarking is a separate dimension focused on measurement rather than the ISA itself.",
                "with_dim_id": "codec-performance-benchmarking",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              },
              {
                "reason": "Both involve hardware-specific optimization, but x86 is a general-purpose CPU architecture rather than an ML accelerator.",
                "with_dim_id": "accelerators-and-hardware-for-ml",
                "with_dim_name": null,
                "with_role": "ML Engineer"
              }
            ],
            "tentative_id": "d_init_01"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "x86",
          "placement_confidence": 0.92,
          "primary_dimension": "d_init_01",
          "reasoning": "Deterministic JD placement: locked_dimensions has 1 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [],
          "skill_id": "x86"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [],
          "related_to": [
            "arm",
            "simd-intrinsics",
            "c",
            "memory-management",
            "memory-profiling",
            "intel-quicksync-video",
            "amd-video-core-next-vcn",
            "nvidia-nvdec"
          ],
          "requires": [],
          "skill_id": "x86",
          "suppress_on_match": []
        },
        "skill_id": "x86",
        "split_log": [],
        "typed": {
          "alternatives_considered": [
            "Concept: ruled out \u2014 although it is a technical term to know, the primary classification is the system-shape/ISA definition rather than a general knowledge concept."
          ],
          "confidence": 0.97,
          "name": "x86",
          "reasoning": "x86 is fundamentally an instruction set architecture, and the Architecture vs Concept rule applies because it describes a system shape for how software is built and executed rather than a knowledge unit.",
          "skill_id": "x86",
          "subtype": "instruction_set_architecture",
          "type": "Architecture"
        },
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "SSE",
          "alias_type": "CANONICAL",
          "id": 2688,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "SSE",
        "id": 1718,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "sse",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Hardware Acceleration and SIMD",
            "id": 235,
            "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
            "slug": "hardware-acceleration-and-simd",
            "source": "db"
          },
          "input_skill": "SSE",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "SSE",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "AVX2",
          "alias_type": "CANONICAL",
          "id": 2686,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "AVX2",
        "id": 1716,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "avx2",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Hardware Acceleration and SIMD",
            "id": 235,
            "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
            "slug": "hardware-acceleration-and-simd",
            "source": "db"
          },
          "input_skill": "AVX",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "AVX",
      "matched_via": "embedding_alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "AArch64",
          "alias_type": "VERSION",
          "id": 2575,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "ARM",
          "alias_type": "CANONICAL",
          "id": 2573,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "ARMv8",
          "alias_type": "VERSION",
          "id": 2574,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "arm64",
          "alias_type": "VERSION",
          "id": 2576,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 1,
        "display_name": "ARM",
        "id": 1621,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PATTERN",
        "slug": "arm",
        "sub_category_id": 1222,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "ARM",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "ARM",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "NEON",
          "alias_type": "CANONICAL",
          "id": 2687,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "NEON",
        "id": 1717,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "neon",
        "sub_category_id": 1277,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Hardware Acceleration and SIMD",
            "id": 235,
            "rationale": "Low-level performance techniques that exploit CPU vector units and hardware decode/encode blocks. This cluster matters because codec work is often constrained by throughput, latency, and power efficiency.",
            "slug": "hardware-acceleration-and-simd",
            "source": "db"
          },
          "input_skill": "NEON",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "NEON",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "Git",
          "alias_type": "CANONICAL",
          "id": 1613,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 13,
        "display_name": "Git",
        "id": 1002,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "git",
        "sub_category_id": 730,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "Git",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "Git",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "AV1",
          "alias_type": "CANONICAL",
          "id": 2648,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 12,
        "display_name": "AV1",
        "id": 1678,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "av1",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Standards and Bitstreams",
            "id": 227,
            "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
            "slug": "codec-standards-and-bitstreams",
            "source": "db"
          },
          "input_skill": "AV1",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "AV1",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "H.265/HEVC",
          "alias_type": "CANONICAL",
          "id": 2647,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 12,
        "display_name": "H.265/HEVC",
        "id": 1677,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "h-265-hevc",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Standards and Bitstreams",
            "id": 227,
            "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
            "slug": "codec-standards-and-bitstreams",
            "source": "db"
          },
          "input_skill": "H.265",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "H.265",
      "matched_via": "embedding_alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "H.264/AVC",
          "alias_type": "CANONICAL",
          "id": 2646,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 12,
        "display_name": "H.264/AVC",
        "id": 1676,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "h-264-avc",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Standards and Bitstreams",
            "id": 227,
            "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
            "slug": "codec-standards-and-bitstreams",
            "source": "db"
          },
          "input_skill": "H.264",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "H.264",
      "matched_via": "embedding_alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "MPEG-2",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "MPEG-2",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Concept",
          "skill_nature": "CONCEPT",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "DEPRECATED"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": false,
            "confused_with": [],
            "reasoning": "MPEG-2 is a specific video compression standard; typical JDs won\u2019t confuse it with other unrelated skills."
          },
          "context_keywords": {
            "context_keywords": [
              "video compression",
              "bitstream",
              "MPEG-2 Transport Stream",
              "MPEG-2 Program Stream",
              "interlaced video",
              "frame rate",
              "encoding",
              "decoding",
              "streaming",
              "digital television",
              "multimedia",
              "video quality",
              "error resilience",
              "bitrate",
              "container format"
            ]
          },
          "maturity": {
            "confidence": 0.93,
            "maturity": "deprecated",
            "reasoning": "MPEG-2 is largely superseded in new deployments by H.264/H.265 and AV1; recent job postings rarely list it except in legacy broadcast/video systems, and modern vendor docs focus on newer codecs."
          },
          "skill_id": "mpeg-2",
          "vendor_license": {
            "confidence": 0.95,
            "license": null,
            "vendor": null,
            "year_introduced": null
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [],
        "locked_dimensions": [
          {
            "description": "Standards and specifications for encoding and decoding digital video streams. MPEG-2 belongs here because it defines a widely used video compression format, bitstream syntax, and interoperability rules for broadcast and disc media.",
            "exemplar_skills": [
              "MPEG-2",
              "MPEG-1",
              "H.262",
              "video compression standards",
              "bitstream syntax",
              "GOP structure",
              "motion compensation"
            ],
            "in_scope": "MPEG-2, MPEG-1, H.262, video bitstream syntax, GOP structure, I/P/B frames, motion compensation, quantization, entropy coding, broadcast video codecs, disc video formats",
            "name": "Video Compression Standards",
            "out_of_scope": "Audio codecs, container formats like MP4 or MKV, streaming protocols, codec benchmarking, defect analysis of compressed artifacts",
            "overlap_flags": [
              {
                "reason": "Codec standards are often evaluated with benchmarking, but this dimension is about the specification itself rather than measurement.",
                "with_dim_id": "codec-performance-benchmarking",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              },
              {
                "reason": "Compression standards can be involved in artifact diagnosis, but that dimension focuses on visible defects and root-cause analysis.",
                "with_dim_id": "artifact-and-defect-analysis",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              }
            ],
            "tentative_id": "d_init_01"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "MPEG-2",
          "placement_confidence": 0.92,
          "primary_dimension": "d_init_01",
          "reasoning": "Deterministic JD placement: locked_dimensions has 1 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [],
          "skill_id": "mpeg-2"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [],
          "related_to": [
            "bitstream-conformance",
            "fragmented-mp4",
            "hardware-decoder-validation",
            "4-2-0",
            "10-bit-color",
            "libavcodec",
            "nvidia-nvdec",
            "intel-quicksync-video"
          ],
          "requires": [],
          "skill_id": "mpeg-2",
          "suppress_on_match": []
        },
        "skill_id": "mpeg-2",
        "split_log": [],
        "typed": {
          "alternatives_considered": [],
          "confidence": 0.0,
          "name": "MPEG-2",
          "reasoning": "Stage 4 failed; fallback typed record.",
          "skill_id": "mpeg-2",
          "subtype": "general",
          "type": "Concept"
        },
        "warnings": [
          "stage4_type_assigner_failed: NotFoundError: Error code: 404 - {\u0027error\u0027: {\u0027code\u0027: \u0027DeploymentNotFound\u0027, \u0027message\u0027: \u0027The API deployment for this resource does not exist. If you created the deployment within the last 5 minutes, please wait a moment and try again.\u0027}}",
          "stage4_used_fallback_typed"
        ]
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "VP9",
          "alias_type": "CANONICAL",
          "id": 2649,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 12,
        "display_name": "VP9",
        "id": 1679,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "STANDARD",
        "slug": "vp9",
        "sub_category_id": 1308,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Standards and Bitstreams",
            "id": 227,
            "rationale": "Formal video codec specifications and the syntax rules that govern interoperability. Engineers use this to ensure encoded output and decoders conform across devices and ecosystems.",
            "slug": "codec-standards-and-bitstreams",
            "source": "db"
          },
          "input_skill": "VP9",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "VP9",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "OpenCL",
          "llm_role": null,
          "roles_from_db": []
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Accelerators and Hardware for ML",
            "id": 58,
            "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
            "slug": "accelerators-and-hardware-for-ml",
            "source": "db"
          },
          "input_skill": "OpenCL",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Accelerators and Hardware for ML",
            "id": 58,
            "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
            "slug": "accelerators-and-hardware-for-ml",
            "source": "db"
          },
          "input_skill": "OpenCL",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "OpenCL",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Language",
          "skill_nature": "LANGUAGE",
          "sub_category": "parallel_computing_api_language",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "STABLE"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": false,
            "confused_with": [],
            "reasoning": "OpenCL is a specific parallel computing API/language; typical JDs won\u2019t confuse it with other common skills in the catalog."
          },
          "context_keywords": {
            "context_keywords": [
              "GPU",
              "parallelism",
              "kernels",
              "cl_device_id",
              "cl_mem",
              "cl_command_queue",
              "cl_program",
              "cl_context",
              "compute units",
              "OpenCL C",
              "platforms",
              "device drivers",
              "memory management",
              "task parallelism",
              "heterogeneous computing"
            ]
          },
          "maturity": {
            "confidence": 0.86,
            "maturity": "niche",
            "reasoning": "OpenCL still appears in GPU/HPC job postings, but far less often than CUDA or vendor SDKs; GitHub activity is steady yet modest, indicating specialized use rather than broad hiring demand."
          },
          "skill_id": "opencl",
          "vendor_license": {
            "confidence": 0.95,
            "license": "other_open",
            "vendor": "Khronos Group",
            "year_introduced": 2008
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [
          {
            "a_dim_id": "d_init_01",
            "a_name": "GPU Compute Programming",
            "a_role": "__skill_focal__",
            "b_dim_id": "accelerators-and-hardware-for-ml",
            "b_name": "Accelerators and Hardware for ML",
            "b_role": "__skill_focal__",
            "pair_kind": "intra_role",
            "reasoning": "Dim A is a programming-model skill cluster: it is about writing parallel compute kernels and managing host-device execution on GPUs/accelerators, with concrete exemplars like OpenCL, OpenCL C, GPU kernel programming, and work-group optimization. Dim B is a hardware-domain cluster: it covers specialized accelerators and hardware for ML training/serving, which is about selecting, understanding, or operating ML chips/systems rather than programming kernels. career-track: no, because a senior OpenCL/GPU kernel programmer is not naturally a senior practitioner in ML accelerator hardware/platform specialization; the daily work, tools, and expertise differ substantially. The similarity is driven by the shared word \u0027accelerators,\u0027 but the substantive skill clusters are distinct.",
            "similarity": 0.670620619855089
          },
          {
            "a_dim_id": "accelerators-and-hardware-for-ml",
            "a_name": "Accelerators and Hardware for ML",
            "a_role": "__skill_focal__",
            "b_dim_id": "accelerators-and-hardware-for-ml",
            "b_name": "Accelerators and Hardware for ML",
            "b_role": "ML Engineer",
            "pair_kind": "cross_role",
            "reasoning": "Dim A is about accelerator programming and hardware-aware compute: OpenCL, GPU acceleration, heterogeneous compute, accelerator kernels, device memory models, and parallel execution on GPUs. Dim B is ML-specific: specialized hardware/accelerators for training and serving machine learning models. career-track: no, because a senior OpenCL/GPU/heterogeneous-compute engineer is not automatically a senior ML model training/serving accelerator engineer; the daily work and expertise differ.",
            "similarity": 0.8185454178924559
          }
        ],
        "locked_dimensions": [
          {
            "description": "Programming models and APIs for writing parallel compute kernels that run on GPUs and other accelerators. OpenCL belongs here because it defines portable kernel execution, memory management, and host-device coordination for heterogeneous compute.",
            "exemplar_skills": [
              "OpenCL",
              "OpenCL C",
              "GPU kernel programming",
              "host-device memory management",
              "work-group optimization",
              "command queue management"
            ],
            "in_scope": "OpenCL, OpenCL C kernels, host-device memory transfers, command queues, buffers, images, kernel compilation, work-item/work-group parallelism, heterogeneous compute programming",
            "name": "GPU Compute Programming",
            "out_of_scope": "Codec algorithm design, video compression standards, CPU-only SIMD tuning, benchmark methodology, driver installation and system setup",
            "overlap_flags": [
              {
                "reason": "Both involve accelerator programming, but this dimension is specifically about general-purpose GPU compute APIs rather than ML accelerator stacks.",
                "with_dim_id": "accelerators-and-hardware-for-ml",
                "with_dim_name": null,
                "with_role": "ML Engineer"
              }
            ],
            "tentative_id": "d_init_01"
          },
          {
            "description": "Specialized hardware and accelerator programming used to offload compute-intensive workloads. OpenCL can fit here when used as a portable GPU/accelerator interface, though the dimension is broader than ML and includes hardware-aware execution.",
            "exemplar_skills": [
              "OpenCL",
              "GPU acceleration",
              "heterogeneous compute",
              "accelerator kernels",
              "device memory management",
              "parallel execution on GPUs"
            ],
            "in_scope": "OpenCL, GPU acceleration, heterogeneous compute, accelerator kernels, device memory models, parallel execution on GPUs, vendor-neutral compute APIs",
            "name": "Accelerators and Hardware for ML",
            "out_of_scope": "ML model training frameworks, distributed training orchestration, codec-specific optimization, graphics rendering APIs, CPU threading libraries",
            "overlap_flags": [
              {
                "reason": "Video codec engineers may use OpenCL for performance work, but benchmarking itself is a separate dimension focused on measurement rather than compute programming.",
                "with_dim_id": "codec-performance-benchmarking",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              }
            ],
            "tentative_id": "accelerators-and-hardware-for-ml"
          },
          {
            "description": "Specialized hardware and accelerators for training and serving machine learning models.",
            "exemplar_skills": [
              "Accelerators and Hardware for ML"
            ],
            "in_scope": "Skills, tools, and practices that belong under Accelerators and Hardware for ML for the target role, including items implied by the dimension rationale.",
            "name": "Accelerators and Hardware for ML",
            "out_of_scope": "Adjacent clusters explicitly not owned by Accelerators and Hardware for ML, including unrelated platforms, roles, and skill families per library policy.",
            "overlap_flags": [],
            "tentative_id": "accelerators-and-hardware-for-ml"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "OpenCL",
          "placement_confidence": 0.92,
          "primary_dimension": "d_init_01",
          "reasoning": "Deterministic JD placement: locked_dimensions has 3 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [
            "accelerators-and-hardware-for-ml"
          ],
          "skill_id": "opencl"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [],
          "related_to": [
            "simd-intrinsics",
            "rendering-efficiency",
            "pytorch",
            "async-programming",
            "competitive-programming",
            "webrtc"
          ],
          "requires": [],
          "skill_id": "opencl",
          "suppress_on_match": []
        },
        "skill_id": "opencl",
        "split_log": [],
        "typed": {
          "alternatives_considered": [],
          "confidence": 0.9,
          "name": "OpenCL",
          "reasoning": "OpenCL is best treated as a programming language/API specification for expressing parallel kernels and host code, so it fits the Language type rather than a tool or framework.",
          "skill_id": "opencl",
          "subtype": "parallel_computing_api_language",
          "type": "Language"
        },
        "warnings": [
          "stage3_post_filter_dropped_catalog_only_locked_dims:42-\u003e3"
        ]
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Accelerators and Hardware for ML",
            "id": 58,
            "rationale": "Specialized hardware and accelerators for training and serving machine learning models.",
            "slug": "accelerators-and-hardware-for-ml",
            "source": "db"
          },
          "input_skill": "CUDA",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "CUDA",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Language",
          "skill_nature": "LANGUAGE",
          "sub_category": "gpu_programming_language",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "STABLE"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": false,
            "confused_with": [],
            "reasoning": "CUDA is a specific NVIDIA GPU programming platform; typical JDs won\u2019t confuse it with other distinct skills in the catalog."
          },
          "context_keywords": {
            "context_keywords": [
              "NVIDIA",
              "GPU",
              "parallel computing",
              "CUDA cores",
              "cuDNN",
              "TensorRT",
              "CUDA Toolkit",
              "kernel",
              "memory management",
              "CUDA streams",
              "thrust",
              "OpenCL",
              "GPGPU",
              "performance optimization",
              "compute capability"
            ]
          },
          "maturity": {
            "confidence": 0.93,
            "maturity": "well_known",
            "reasoning": "CUDA appears in many ML/HPC job descriptions and is the de facto NVIDIA GPU programming stack; NVIDIA continues active platform support and ecosystem investment, indicating broad market adoption."
          },
          "skill_id": "cuda",
          "vendor_license": {
            "confidence": 0.95,
            "license": "proprietary",
            "vendor": "NVIDIA",
            "year_introduced": 2006
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [],
        "locked_dimensions": [
          {
            "description": "Programming and optimization work that targets GPUs and other accelerators for high-throughput compute. CUDA belongs here because it is the primary programming model for NVIDIA GPU execution, memory movement, and kernel optimization.",
            "exemplar_skills": [
              "CUDA",
              "CUDA kernels",
              "GPU memory management",
              "shared memory optimization",
              "warp-level programming",
              "kernel launch tuning"
            ],
            "in_scope": "CUDA, CUDA kernels, GPU memory hierarchy, shared memory, warp-level primitives, streams and events, kernel launch configuration, occupancy tuning, NVIDIA GPU programming",
            "name": "GPU Accelerators and Parallel Compute",
            "out_of_scope": "GPU cluster scheduling, distributed training frameworks, model parallel orchestration, CPU-only optimization, cloud instance selection, video codec algorithm design",
            "overlap_flags": [
              {
                "reason": "CUDA is often used inside distributed training stacks, but this dimension owns the multi-GPU orchestration layer rather than low-level kernel programming.",
                "with_dim_id": "distributed-training-systems",
                "with_dim_name": null,
                "with_role": "ML Engineer, ML Ops Engineer"
              },
              {
                "reason": "CUDA may be used to accelerate codec workloads, but benchmarking and profiling of codec implementations is a separate concern.",
                "with_dim_id": "codec-performance-benchmarking",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              }
            ],
            "tentative_id": "accelerators-and-hardware-for-ml"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "CUDA",
          "placement_confidence": 0.92,
          "primary_dimension": "accelerators-and-hardware-for-ml",
          "reasoning": "Deterministic JD placement: locked_dimensions has 1 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [],
          "skill_id": "cuda"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [
            "c",
            "python",
            "simd-intrinsics"
          ],
          "related_to": [
            "nvidia-nvdec",
            "nvidia-gpu-operator",
            "tensorflow",
            "pytorch-ddp",
            "rendering-efficiency",
            "amd-video-core-next-vcn",
            "arm",
            "api"
          ],
          "requires": [],
          "skill_id": "cuda",
          "suppress_on_match": []
        },
        "skill_id": "cuda",
        "split_log": [],
        "typed": {
          "alternatives_considered": [],
          "confidence": 0.9,
          "name": "CUDA",
          "reasoning": "CUDA is fundamentally a programming language/toolchain for expressing GPU kernels and device code, so it fits the Language category rather than a library or framework.",
          "skill_id": "cuda",
          "subtype": "gpu_programming_language",
          "type": "Language"
        },
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "Machine Learning",
          "alias_type": "CANONICAL",
          "id": 2015,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "Machine Learning",
        "id": 1356,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "machine-learning",
        "sub_category_id": 1024,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "AI Governance and Model Security",
            "id": 50,
            "rationale": "Controls and documentation used to make models safer, auditable, and compliant. ML engineers use this to manage model risk, supply chain integrity, and governance requirements.",
            "slug": "ai-governance-and-model-security",
            "source": "db"
          },
          "input_skill": "Machine Learning",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            },
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            },
            {
              "display_name": "ML Ops Engineer",
              "id": 16,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-ops-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "Machine Learning",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "Machine Learning",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "Neural Networks",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "Neural Networks",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Concept",
          "skill_nature": "CONCEPT",
          "sub_category": "machine_learning_model_concept",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "STABLE"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": false,
            "confused_with": [],
            "reasoning": "\u201cNeural Networks\u201d is a specific ML model concept; typical JDs won\u2019t confuse it with other distinct skills in the catalog."
          },
          "context_keywords": {
            "context_keywords": [
              "backpropagation",
              "activation functions",
              "convolutional layers",
              "recurrent networks",
              "dropout",
              "gradient descent",
              "TensorFlow",
              "Keras",
              "PyTorch",
              "overfitting",
              "training data",
              "hyperparameters",
              "transfer learning",
              "model architecture",
              "regularization"
            ]
          },
          "maturity": {
            "confidence": 0.98,
            "maturity": "well_known",
            "reasoning": "Broadly listed in ML/AI job descriptions and core to modern deep learning stacks; major frameworks like PyTorch and TensorFlow center on neural networks."
          },
          "skill_id": "neural-networks",
          "vendor_license": {
            "confidence": 0.95,
            "license": null,
            "vendor": null,
            "year_introduced": null
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [],
        "locked_dimensions": [
          {
            "description": "Core concepts, architectures, and training behavior of artificial neural networks. This fits the target skill because it refers to the model family itself rather than a specific deployment, hardware, or operations stack.",
            "exemplar_skills": [
              "Neural Networks",
              "Backpropagation",
              "Activation Functions",
              "Multilayer Perceptrons",
              "Convolutional Neural Networks",
              "Recurrent Neural Networks"
            ],
            "in_scope": "Neural Networks, perceptrons, multilayer perceptrons, backpropagation, activation functions, loss functions, optimization basics, feedforward networks, recurrent neural networks, convolutional neural networks, attention mechanisms, regularization",
            "name": "Neural Network Fundamentals",
            "out_of_scope": "Distributed training frameworks and multi-GPU scaling, model serving infrastructure, accelerator hardware selection, codec-specific implementation details, general machine learning lifecycle operations",
            "overlap_flags": [
              {
                "reason": "Neural network training can be scaled across devices, but that dimension owns the distributed execution mechanics rather than the model concept itself.",
                "with_dim_id": "distributed-training-systems",
                "with_dim_name": null,
                "with_role": "ML Engineer, ML Ops Engineer"
              },
              {
                "reason": "Neural networks often run on specialized hardware, but hardware selection and optimization are separate from the model architecture skill.",
                "with_dim_id": "accelerators-and-hardware-for-ml",
                "with_dim_name": null,
                "with_role": "ML Engineer"
              }
            ],
            "tentative_id": "d_init_01"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "Neural Networks",
          "placement_confidence": 0.92,
          "primary_dimension": "d_init_01",
          "reasoning": "Deterministic JD placement: locked_dimensions has 1 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [],
          "skill_id": "neural-networks"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [],
          "related_to": [
            "scikit-learn",
            "tensorflow",
            "pytorch",
            "vision-language-models",
            "multimodal-document-understanding",
            "anomaly-detection",
            "hybrid-retrieval",
            "agentic-systems"
          ],
          "requires": [],
          "skill_id": "neural-networks",
          "suppress_on_match": []
        },
        "skill_id": "neural-networks",
        "split_log": [],
        "typed": {
          "alternatives_considered": [],
          "confidence": 0.97,
          "name": "Neural Networks",
          "reasoning": "Neural Networks are a named knowledge unit in machine learning, so by the Concept vs Methodology rule they are a Concept rather than an Architecture or Framework.",
          "skill_id": "neural-networks",
          "subtype": "machine_learning_model_concept",
          "type": "Concept"
        },
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "multithreading",
          "alias_type": "CANONICAL",
          "id": 223,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "multithreading",
        "id": 82,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "multithreading",
        "sub_category_id": 7,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Concurrency and Parallel Processing",
            "id": 17,
            "rationale": "Programming techniques for handling multiple requests and background work safely and efficiently. Includes synchronization, async execution, and coordination of concurrent tasks.",
            "slug": "concurrency-and-parallel-processing",
            "source": "db"
          },
          "input_skill": "Multithreading",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Backend Engineer",
              "id": 1,
              "rationale": null,
              "role_archetype": "A Backend Engineer designs, builds, and maintains the server-side logic and data handling that power applications and services. They focus on implementing reliable business functionality, integrating with other systems, and ensuring the backend is scalable, maintainable, and observable.",
              "slug": "backend-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Multithreading",
      "matched_via": "alias",
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": null,
      "source_tag": "db",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Performance Benchmarking",
            "id": 238,
            "rationale": "Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.",
            "slug": "codec-performance-benchmarking",
            "source": "db"
          },
          "input_skill": "Cache Optimization",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "React Frontend Development",
            "id": 96,
            "rationale": "Building interactive web user interfaces with React.js, including component composition, state management, hooks, and rendering patterns. React.js belongs here because it is a core library for client-side UI development in modern web applications.",
            "slug": "d_init_01",
            "source": "db"
          },
          "input_skill": "Cache Optimization",
          "llm_role": null,
          "roles_from_db": []
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Codec Performance Benchmarking",
            "id": 238,
            "rationale": "Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.",
            "slug": "codec-performance-benchmarking",
            "source": "db"
          },
          "input_skill": "Cache Optimization",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Video Codec Engineer",
              "id": 22,
              "rationale": null,
              "role_archetype": null,
              "slug": "video-codec-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Cache Optimization",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Concept",
          "skill_nature": "CONCEPT",
          "sub_category": "cache_optimization",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "NOT_APPLICABLE",
          "volatility": "STABLE"
        },
        "enrichment": {
          "ambiguity": {
            "ambiguity_flag": false,
            "confused_with": [],
            "reasoning": "\u201cCache Optimization\u201d is a specific performance-tuning concept; it\u2019s unlikely to be confused with other distinct catalog skills in typical JDs."
          },
          "context_keywords": {
            "context_keywords": [
              "cache eviction",
              "cache hit ratio",
              "cache coherence",
              "LRU",
              "memcached",
              "Redis",
              "CDN",
              "data locality",
              "prefetching",
              "write-through",
              "write-back",
              "cache partitioning",
              "distributed caching",
              "cache warming",
              "performance tuning"
            ]
          },
          "maturity": {
            "confidence": 0.84,
            "maturity": "well_known",
            "reasoning": "Common performance topic in JDs for backend, systems, and mobile roles; cache-miss reduction and CPU cache locality are standard interview and profiling concerns across major stacks."
          },
          "skill_id": "cache-optimization",
          "vendor_license": {
            "confidence": 0.95,
            "license": null,
            "vendor": null,
            "year_introduced": null
          },
          "versioning": {
            "current_version": null,
            "version_aliases": {},
            "versioned": false
          }
        },
        "keep_log": [],
        "locked_dimensions": [
          {
            "description": "Measurement and profiling practices used to improve codec speed, memory use, and throughput. Cache optimization belongs here because it is a core performance-tuning technique for reducing stalls and improving encode/decode efficiency.",
            "exemplar_skills": [
              "Cache Optimization",
              "Performance Profiling",
              "Cache Miss Analysis",
              "Memory Locality Tuning",
              "Codec Benchmarking",
              "Throughput Optimization"
            ],
            "in_scope": "Cache Optimization, profiling cache misses, cache-aware data layout, memory locality tuning, SIMD-friendly access patterns, encoder/decoder throughput measurement, latency and throughput benchmarking, CPU cache behavior analysis",
            "name": "Codec Performance Benchmarking",
            "out_of_scope": "Compression algorithm design, artifact quality analysis, hardware driver integration, general application performance tuning, distributed systems scaling, GPU kernel optimization",
            "overlap_flags": [
              {
                "reason": "Cache behavior can indirectly affect visible codec artifacts, but that dimension focuses on defect diagnosis rather than performance tuning.",
                "with_dim_id": "artifact-and-defect-analysis",
                "with_dim_name": null,
                "with_role": "Video Codec Engineer"
              }
            ],
            "tentative_id": "codec-performance-benchmarking"
          },
          {
            "description": "Techniques for improving how software uses CPU caches and memory hierarchy. Cache Optimization fits here because it specifically targets locality, reduced misses, and better access patterns in performance-critical systems.",
            "exemplar_skills": [
              "Cache Optimization",
              "Memory Locality Tuning",
              "Cache Miss Reduction",
              "Data Layout Optimization",
              "Spatial Locality",
              "Temporal Locality"
            ],
            "in_scope": "Cache Optimization, cache-aware algorithms, data structure layout, spatial locality, temporal locality, prefetch-friendly access patterns, false sharing reduction, memory access profiling",
            "name": "Memory Locality Optimization",
            "out_of_scope": "General benchmarking methodology, codec-specific quality defects, GPU memory management, operating system virtual memory policy, distributed cache systems",
            "overlap_flags": [
              {
                "reason": "Both involve improving runtime efficiency, but this dimension is narrower and centered on CPU cache and memory access behavior.",
                "with_dim_id": "performance-and-stability-tuning",
                "with_dim_name": null,
                "with_role": "Android Engineer, Ios engineer"
              }
            ],
            "tentative_id": "d_init_01"
          },
          {
            "description": "Measurement and profiling practices used to compare codec implementations and tune resource usage. Engineers rely on this to quantify speed, memory, and quality tradeoffs across builds and platforms.",
            "exemplar_skills": [
              "Codec Performance Benchmarking"
            ],
            "in_scope": "Skills, tools, and practices that belong under Codec Performance Benchmarking for the target role, including items implied by the dimension rationale.",
            "name": "Codec Performance Benchmarking",
            "out_of_scope": "Adjacent clusters explicitly not owned by Codec Performance Benchmarking, including unrelated platforms, roles, and skill families per library policy.",
            "overlap_flags": [],
            "tentative_id": "codec-performance-benchmarking"
          }
        ],
        "merge_log": [],
        "placed": {
          "name": "Cache Optimization",
          "placement_confidence": 0.92,
          "primary_dimension": "codec-performance-benchmarking",
          "reasoning": "Deterministic JD placement: locked_dimensions has 3 dimension(s) from skill-driven dimension generation after reconciliation; primary_dimension is the first locked dim.",
          "secondary_dimensions": [
            "d_init_01"
          ],
          "skill_id": "cache-optimization"
        },
        "relationships": {
          "child_skills": [],
          "parent_skills": [],
          "related_to": [
            "image-caching",
            "rendering-efficiency",
            "memory-profiling",
            "memory-management",
            "offline-first",
            "hybrid-retrieval",
            "async-programming",
            "async-processing"
          ],
          "requires": [],
          "skill_id": "cache-optimization",
          "suppress_on_match": []
        },
        "skill_id": "cache-optimization",
        "split_log": [],
        "typed": {
          "alternatives_considered": [],
          "confidence": 0.93,
          "name": "Cache Optimization",
          "reasoning": "This is best treated as a Concept because it names a technical knowledge unit about improving cache behavior, not a specific tool, methodology, or architecture.",
          "skill_id": "cache-optimization",
          "subtype": "cache_optimization",
          "type": "Concept"
        },
        "warnings": [
          "stage3_post_filter_dropped_catalog_only_locked_dims:42-\u003e3"
        ]
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    }
  ],
  "unmatched_skills": [
    "x86",
    "MPEG-2",
    "OpenCL",
    "CUDA",
    "Neural Networks",
    "Cache Optimization"
  ]
}

API 3 — final-role-output

{}

LLM Calls

Every model call made for this run, in pipeline order. Click a card to see the model's response.

Loading…