Pipeline run

{
  "final_skills": [
    {
      "is_primary": true,
      "skill_name": "Python"
    },
    {
      "is_primary": true,
      "skill_name": "PyTorch"
    },
    {
      "is_primary": false,
      "skill_name": "TensorFlow"
    },
    {
      "is_primary": false,
      "skill_name": "JAX"
    },
    {
      "is_primary": true,
      "skill_name": "Hugging Face Transformers"
    },
    {
      "is_primary": true,
      "skill_name": "Llama 3"
    },
    {
      "is_primary": true,
      "skill_name": "Mistral"
    },
    {
      "is_primary": true,
      "skill_name": "RAG"
    },
    {
      "is_primary": true,
      "skill_name": "FAISS"
    },
    {
      "is_primary": true,
      "skill_name": "pgvector"
    },
    {
      "is_primary": true,
      "skill_name": "A/B Testing"
    },
    {
      "is_primary": true,
      "skill_name": "MLflow"
    },
    {
      "is_primary": true,
      "skill_name": "CI/CD"
    },
    {
      "is_primary": false,
      "skill_name": "Pinecone"
    },
    {
      "is_primary": false,
      "skill_name": "Weaviate"
    },
    {
      "is_primary": false,
      "skill_name": "text-embedding-3"
    },
    {
      "is_primary": false,
      "skill_name": "BGE"
    },
    {
      "is_primary": false,
      "skill_name": "vLLM"
    },
    {
      "is_primary": false,
      "skill_name": "TGI"
    },
    {
      "is_primary": false,
      "skill_name": "Triton"
    },
    {
      "is_primary": false,
      "skill_name": "Kubeflow"
    },
    {
      "is_primary": false,
      "skill_name": "Weights \u0026 Biases"
    },
    {
      "is_primary": false,
      "skill_name": "Airflow"
    },
    {
      "is_primary": false,
      "skill_name": "Prefect"
    },
    {
      "is_primary": false,
      "skill_name": "Kubernetes"
    },
    {
      "is_primary": false,
      "skill_name": "GPU scheduling"
    },
    {
      "is_primary": false,
      "skill_name": "BM25"
    },
    {
      "is_primary": false,
      "skill_name": "Quantization"
    },
    {
      "is_primary": false,
      "skill_name": "FP16"
    },
    {
      "is_primary": false,
      "skill_name": "INT8"
    }
  ],
  "jd_role": {
    "display_name": "Senior Machine Learning Engineer",
    "rationale": null,
    "role_aliases": [
      "ML Engineer",
      "Machine Learning Engineer",
      "Senior ML Engineer"
    ],
    "role_archetype": "Data",
    "slug": ""
  },
  "nano_parsed": {
    "JD_type": "pass",
    "about_company": null,
    "certifications": [],
    "company_name": null,
    "ctc": {
      "currency": null,
      "max": null,
      "min": null,
      "period": null,
      "raw": "competitive"
    },
    "domain": {
      "primary": {
        "aliases": [],
        "domain": "Software \u0026 SaaS Products"
      },
      "secondary": null
    },
    "education": [],
    "experience": {
      "max": null,
      "min": 5,
      "raw": "5+ years of hands-on ML engineering experience, with at least 2 years productionising deep-learning models."
    },
    "job_locations": [
      {
        "aliases": [
          "Bangalore"
        ],
        "city": "Bengaluru",
        "country": "India",
        "state": null,
        "work_mode": "hybrid"
      }
    ],
    "role": "Senior Machine Learning Engineer",
    "role_aliases": [
      "ML Engineer",
      "Machine Learning Engineer",
      "Senior ML Engineer"
    ],
    "role_archetype": "Data",
    "roles_and_responsibilities": [
      {
        "bullet_count": 0,
        "heading": "About the role",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "We are looking for a",
          "last_5_words": "infrastructure end-to-end."
        },
        "text": "We are looking for a Senior ML Engineer to own the lifecycle of production machine-learning systems at our platform. You will work with applied scientists to take research models from notebook to production, build retrieval-augmented generation (RAG) pipelines on top of our internal document corpus, and own MLOps infrastructure end-to-end.",
        "word_count": 49
      },
      {
        "bullet_count": 5,
        "heading": "Responsibilities",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "Design and deploy production-grade",
          "last_5_words": "versioning (MLflow), and CI/CD for ML."
        },
        "text": "- Design and deploy production-grade ML services serving 50M+ predictions/day.\n- Build feature stores, training pipelines, and online inference infrastructure with PyTorch and Hugging Face Transformers.\n- Lead our LLM efforts: fine-tune open-source models (Llama 3, Mistral) and integrate with our RAG retrieval layer (FAISS, pgvector embeddings).\n- Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.\n- Mentor junior ML engineers on best practices for reproducible experiments, model versioning (MLflow), and CI/CD for ML.",
        "word_count": 104
      },
      {
        "bullet_count": 6,
        "heading": "Requirements",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "5+ years of hands-on ML",
          "last_5_words": "experience with GPU scheduling preferred."
        },
        "text": "- 5+ years of hands-on ML engineering experience, with at least 2 years productionising deep-learning models.\n- Strong Python; production code with PyTorch, TensorFlow, or JAX.\n- Experience with vector databases (Pinecone, Weaviate, pgvector) and embedding models (text-embedding-3, BGE, etc.).\n- LLM serving experience: vLLM, TGI, or Triton.\n- Familiar with MLOps tooling (MLflow, Kubeflow, Weights \u0026 Biases) and orchestration (Airflow, Prefect).\n- Comfort with Kubernetes for ML workloads; experience with GPU scheduling preferred.",
        "word_count": 104
      },
      {
        "bullet_count": 3,
        "heading": "Nice to have",
        "heading_was_present": true,
        "source_marker": {
          "first_5_words": "Open-source contributions to ML",
          "last_5_words": "quantization (INT8/FP16), and inference optimisation."
        },
        "text": "- Open-source contributions to ML frameworks or popular LLM tooling.\n- Experience with retrieval-augmented generation, sparse retrieval (BM25), and hybrid search.\n- Familiarity with vector index tuning, quantization (INT8/FP16), and inference optimisation.",
        "word_count": 45
      }
    ],
    "urls": []
  },
  "rejected": false,
  "rejection_reason": null,
  "run_id": "0aa94386-7f59-430c-b6f7-8f25e5b5c76e",
  "stage3_signals": {
    "alias_found": true,
    "alias_match_roles": [
      {
        "display_name": "ML Engineer",
        "kra_matches": null,
        "matched_count": null,
        "matched_skills": null,
        "role_id": 3,
        "score": 1.0,
        "slug": "ml-engineer",
        "total_count": null
      }
    ],
    "kra_match_roles": [
      {
        "display_name": "ML Ops Engineer",
        "kra_matches": [
          {
            "kra_text": "Maintains model versioning, experiment lineage, and artifact tracking using MLflow, DVC, or Weights \u0026 Biases for reproducibility and auditability.",
            "sentence": "Mentor junior ML engineers on best practices for reproducible experiments, model versioning (MLflow), and CI/CD for ML.",
            "similarity": 0.6578
          },
          {
            "kra_text": "Sets up model monitoring dashboards, data drift detection, prediction performance tracking, and alert routing for production ML systems.",
            "sentence": "Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.",
            "similarity": 0.6212
          },
          {
            "kra_text": "Sets up model monitoring dashboards, data drift detection, prediction performance tracking, and alert routing for production ML systems.",
            "sentence": "Design and deploy production-grade ML services serving 50M+ predictions/day.",
            "similarity": 0.6049
          }
        ],
        "matched_count": null,
        "matched_skills": null,
        "role_id": 16,
        "score": 0.628,
        "slug": "ml-ops-engineer",
        "total_count": null
      },
      {
        "display_name": "ML Engineer",
        "kra_matches": [
          {
            "kra_text": "Builds model serving infrastructure to deploy trained models as real-time prediction APIs or batch inference jobs using TorchServe, TensorFlow Serving, or SageMaker.",
            "sentence": "Design and deploy production-grade ML services serving 50M+ predictions/day.",
            "similarity": 0.6091
          },
          {
            "kra_text": "Prepares, cleans, and transforms training datasets, manages feature stores, and builds feature engineering pipelines for model training.",
            "sentence": "Build feature stores, training pipelines, and online inference infrastructure with PyTorch and Hugging Face Transformers.",
            "similarity": 0.607
          },
          {
            "kra_text": "Monitors production model behavior for data drift, concept drift, and prediction performance degradation using monitoring dashboards and alerting.",
            "sentence": "Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.",
            "similarity": 0.5963
          }
        ],
        "matched_count": null,
        "matched_skills": null,
        "role_id": 3,
        "score": 0.6042,
        "slug": "ml-engineer",
        "total_count": null
      },
      {
        "display_name": "AI Engineer",
        "kra_matches": [
          {
            "kra_text": "Defines evaluation frameworks, automated test suites, and human feedback loops to measure AI feature quality, accuracy, and consistency.",
            "sentence": "Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.",
            "similarity": 0.5699
          },
          {
            "kra_text": "Designs and implements prompt engineering workflows, few-shot examples, chain-of-thought patterns, and structured output parsing for AI feature pipelines.",
            "sentence": "Build feature stores, training pipelines, and online inference infrastructure with PyTorch and Hugging Face Transformers.",
            "similarity": 0.5055
          },
          {
            "kra_text": "Designs and implements prompt engineering workflows, few-shot examples, chain-of-thought patterns, and structured output parsing for AI feature pipelines.",
            "sentence": "You will work with applied scientists to take research models from notebook to production, build retrieval-augmented generation (RAG) pipelines on top of our internal document corpus, and own MLOps infrastructure end-to-end.",
            "similarity": 0.4857
          }
        ],
        "matched_count": null,
        "matched_skills": null,
        "role_id": 13,
        "score": 0.5204,
        "slug": "ai-engineer",
        "total_count": null
      },
      {
        "display_name": "DevOps Engineer",
        "kra_matches": [
          {
            "kra_text": "Collaborates with development teams to improve build processes, reduce deployment friction, containerize applications, and adopt DevOps best practices.",
            "sentence": "Mentor junior ML engineers on best practices for reproducible experiments, model versioning (MLflow), and CI/CD for ML.",
            "similarity": 0.4746
          },
          {
            "kra_text": "Manages container orchestration with Kubernetes and Docker, deploying applications as pods, managing namespaces, and configuring auto-scaling across cloud environments.",
            "sentence": "Comfort with Kubernetes for ML workloads; experience with GPU scheduling preferred.",
            "similarity": 0.459
          },
          {
            "kra_text": "Monitors CI/CD pipeline reliability, identifies bottlenecks in delivery workflows, and improves deployment frequency, lead time, and failure recovery rate.",
            "sentence": "Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.",
            "similarity": 0.4484
          }
        ],
        "matched_count": null,
        "matched_skills": null,
        "role_id": 10,
        "score": 0.4607,
        "slug": "devops-engineer",
        "total_count": null
      },
      {
        "display_name": "AI Compliance Officer",
        "kra_matches": [
          {
            "kra_text": "Monitors deployed AI systems for compliance policy drift, regulatory changes, and emerging requirements affecting existing AI deployments.",
            "sentence": "Set up evaluation harnesses, A/B testing frameworks, and model monitoring/drift detection.",
            "similarity": 0.5023
          },
          {
            "kra_text": "Reviews AI use cases and model deployments against applicable regulations, internal ethics policies, and governance guidelines prior to production approval.",
            "sentence": "Design and deploy production-grade ML services serving 50M+ predictions/day.",
            "similarity": 0.4208
          },
          {
            "kra_text": "Monitors deployed AI systems for compliance policy drift, regulatory changes, and emerging requirements affecting existing AI deployments.",
            "sentence": "We are looking for a Senior ML Engineer to own the lifecycle of production machine-learning systems at our platform.",
            "similarity": 0.4038
          }
        ],
        "matched_count": null,
        "matched_skills": null,
        "role_id": 12,
        "score": 0.4423,
        "slug": "ai-compliance-officer",
        "total_count": null
      }
    ],
    "skill_match_roles": [
      {
        "display_name": "ML Engineer",
        "kra_matches": null,
        "matched_count": 5,
        "matched_skills": [
          "CI/CD",
          "FAISS",
          "MLflow",
          "PyTorch",
          "Python"
        ],
        "role_id": 3,
        "score": 0.4545,
        "slug": "ml-engineer",
        "total_count": 11
      },
      {
        "display_name": "ML Ops Engineer",
        "kra_matches": null,
        "matched_count": 4,
        "matched_skills": [
          "FAISS",
          "MLflow",
          "PyTorch",
          "Python"
        ],
        "role_id": 16,
        "score": 0.3636,
        "slug": "ml-ops-engineer",
        "total_count": 11
      },
      {
        "display_name": "AI Engineer",
        "kra_matches": null,
        "matched_count": 2,
        "matched_skills": [
          "PyTorch",
          "pgvector"
        ],
        "role_id": 13,
        "score": 0.1818,
        "slug": "ai-engineer",
        "total_count": 11
      },
      {
        "display_name": "AR/VR Engineer",
        "kra_matches": null,
        "matched_count": 1,
        "matched_skills": [
          "Python"
        ],
        "role_id": 8,
        "score": 0.0909,
        "slug": "ar-vr-engineer",
        "total_count": 11
      },
      {
        "display_name": "DevOps Engineer",
        "kra_matches": null,
        "matched_count": 1,
        "matched_skills": [
          "CI/CD"
        ],
        "role_id": 10,
        "score": 0.0909,
        "slug": "devops-engineer",
        "total_count": 11
      }
    ]
  },
  "stage4_decision": {
    "alias_collision_detected": false,
    "case": "A",
    "chosen_role": {
      "display_name": "ML Engineer",
      "kra_matches": null,
      "matched_count": null,
      "matched_skills": null,
      "role_id": 3,
      "score": 1.0,
      "slug": "ml-engineer",
      "total_count": null
    },
    "confidence": 1.0,
    "is_new_role": false,
    "llm2_fired": false,
    "llm2_reasoning": null,
    "matched_dimensions": [],
    "matched_kras": [],
    "matched_skills": [],
    "new_role_display_name": null,
    "new_role_slug": null,
    "queued": false,
    "reasoning": "Exact alias hit on ml-engineer (1.0) \u2014 no other alias at this confidence; skill_top ml-engineer 0.45 does not contradict",
    "sub_role": null
  },
  "stage5_updates": {
    "centroid_n_after": 15,
    "centroid_updated": true,
    "collision_log_id": null,
    "new_kra_attached": null,
    "new_skills_attached": [
      {
        "is_primary": true,
        "queue_id": 3234,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "Hugging Face Transformers",
        "status": "pending"
      },
      {
        "is_primary": true,
        "queue_id": 3235,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "Llama 3",
        "status": "pending"
      },
      {
        "is_primary": true,
        "queue_id": 3236,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "Mistral",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3237,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "text-embedding-3",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3238,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "TGI",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3239,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "GPU scheduling",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3240,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "BM25",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3241,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "FP16",
        "status": "pending"
      },
      {
        "is_primary": false,
        "queue_id": 3242,
        "role_display_name": "ML Engineer",
        "role_slug": "ml-engineer",
        "skill_name": "INT8",
        "status": "pending"
      }
    ],
    "queue_entry_id": null,
    "v3_pipeline_triggered": false,
    "v3_role_slug": null,
    "v3_run_id": null
  }
}

{
  "alias_matches": [
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 67,
      "existing_alias_text": "Python",
      "input_term": "Python",
      "matched_canonical": {
        "category_id": 6,
        "display_name": "Python",
        "id": 5,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "python",
        "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": 441,
      "existing_alias_text": "PyTorch",
      "input_term": "PyTorch",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "PyTorch",
        "id": 195,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "pytorch",
        "sub_category_id": 156,
        "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": 442,
      "existing_alias_text": "TensorFlow",
      "input_term": "TensorFlow",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "TensorFlow",
        "id": 196,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "tensorflow",
        "sub_category_id": 156,
        "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": 451,
      "existing_alias_text": "JAX",
      "input_term": "JAX",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "JAX",
        "id": 199,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "jax",
        "sub_category_id": 156,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1830,
      "existing_alias_text": "RAG",
      "input_term": "RAG",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "RAG",
        "id": 1194,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "rag",
        "sub_category_id": 904,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1820,
      "existing_alias_text": "FAISS",
      "input_term": "FAISS",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "FAISS",
        "id": 1184,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "faiss",
        "sub_category_id": 894,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1882,
      "existing_alias_text": "pgvector",
      "input_term": "pgvector",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "pgvector",
        "id": 1246,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "pgvector",
        "sub_category_id": 972,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2565,
      "existing_alias_text": "A/B Testing",
      "input_term": "A/B Testing",
      "matched_canonical": {
        "category_id": 8,
        "display_name": "A/B Testing",
        "id": 1613,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "METHODOLOGY",
        "slug": "a-b-testing",
        "sub_category_id": 1214,
        "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": 470,
      "existing_alias_text": "MLflow",
      "input_term": "MLflow",
      "matched_canonical": {
        "category_id": 13,
        "display_name": "MLflow",
        "id": 214,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "mlflow",
        "sub_category_id": 1168,
        "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": 1826,
      "existing_alias_text": "CI/CD",
      "input_term": "CI/CD",
      "matched_canonical": {
        "category_id": 8,
        "display_name": "CI/CD",
        "id": 1190,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "METHODOLOGY",
        "slug": "ci-cd",
        "sub_category_id": 900,
        "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": 503,
      "existing_alias_text": "Pinecone",
      "input_term": "Pinecone",
      "matched_canonical": {
        "category_id": 9,
        "display_name": "Pinecone",
        "id": 242,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "pinecone",
        "sub_category_id": 177,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1878,
      "existing_alias_text": "Weaviate",
      "input_term": "Weaviate",
      "matched_canonical": {
        "category_id": 9,
        "display_name": "Weaviate",
        "id": 1242,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "weaviate",
        "sub_category_id": 177,
        "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": 1888,
      "existing_alias_text": "text-embedding-3-large",
      "input_term": "text-embedding-3",
      "matched_canonical": {
        "category_id": 11,
        "display_name": "text-embedding-3-large",
        "id": 1252,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CLOUD_SERVICE",
        "slug": "text-embedding-3-large",
        "sub_category_id": 1005,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "embedding_alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1885,
      "existing_alias_text": "BGE",
      "input_term": "BGE",
      "matched_canonical": {
        "category_id": 7,
        "display_name": "BGE",
        "id": 1249,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "bge",
        "sub_category_id": 973,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 502,
      "existing_alias_text": "vLLM",
      "input_term": "vLLM",
      "matched_canonical": {
        "category_id": 13,
        "display_name": "vLLM",
        "id": 241,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "vllm",
        "sub_category_id": 1167,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 2026,
      "existing_alias_text": "Triton",
      "input_term": "Triton",
      "matched_canonical": {
        "category_id": 6,
        "display_name": "Triton",
        "id": 1367,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "triton",
        "sub_category_id": 1033,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 465,
      "existing_alias_text": "Kubeflow",
      "input_term": "Kubeflow",
      "matched_canonical": {
        "category_id": 5,
        "display_name": "Kubeflow",
        "id": 213,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "FRAMEWORK",
        "slug": "kubeflow",
        "sub_category_id": 1127,
        "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": 471,
      "existing_alias_text": "Weights \u0026 Biases",
      "input_term": "Weights \u0026 Biases",
      "matched_canonical": {
        "category_id": 9,
        "display_name": "Weights \u0026 Biases",
        "id": 215,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "weights-biases",
        "sub_category_id": 171,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 526,
      "existing_alias_text": "Airflow",
      "input_term": "Airflow",
      "matched_canonical": {
        "category_id": 13,
        "display_name": "Airflow",
        "id": 265,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "airflow",
        "sub_category_id": 130,
        "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": 306,
      "existing_alias_text": "Prefect",
      "input_term": "Prefect",
      "matched_canonical": {
        "category_id": 13,
        "display_name": "Prefect",
        "id": 112,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "prefect",
        "sub_category_id": 130,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "matched_via": "alias"
    },
    {
      "alias_persist_skipped_reason": "alias_text already exists for this canonical skill",
      "alias_persisted": false,
      "existing_alias_id": 1267,
      "existing_alias_text": "Kubernetes",
      "input_term": "Kubernetes",
      "matched_canonical": {
        "category_id": 9,
        "display_name": "Kubernetes",
        "id": 726,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "kubernetes",
        "sub_category_id": 557,
        "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": 521,
      "existing_alias_text": "quantization",
      "input_term": "Quantization",
      "matched_canonical": {
        "category_id": 2,
        "display_name": "quantization",
        "id": 260,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "quantization",
        "sub_category_id": 1282,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "matched_via": "alias"
    }
  ],
  "candidate_roles": [
    {
      "display_name": "Cloud Security Engineer",
      "id": 23,
      "rationale": null,
      "role_archetype": null,
      "slug": "cloud-security-engineer",
      "source": "db"
    },
    {
      "display_name": "Backend Developer",
      "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": "Fullstack Developer",
      "id": 15,
      "rationale": null,
      "role_archetype": null,
      "slug": "full-stack-engineer",
      "source": "db"
    },
    {
      "display_name": "Cybersecurity Engineer",
      "id": 5,
      "rationale": null,
      "role_archetype": null,
      "slug": "cybersecurity-engineer",
      "source": "db"
    },
    {
      "display_name": "Data Engineer",
      "id": 2,
      "rationale": null,
      "role_archetype": null,
      "slug": "data-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": "AI Engineer",
      "id": 13,
      "rationale": null,
      "role_archetype": null,
      "slug": "ai-engineer",
      "source": "db"
    },
    {
      "display_name": "DevOps Engineer",
      "id": 10,
      "rationale": null,
      "role_archetype": null,
      "slug": "devops-engineer",
      "source": "db"
    },
    {
      "display_name": "Cloud Architect",
      "id": 9,
      "rationale": null,
      "role_archetype": null,
      "slug": "cloud-architect",
      "source": "db"
    },
    {
      "display_name": "Video Codec Engineer",
      "id": 22,
      "rationale": null,
      "role_archetype": null,
      "slug": "video-codec-engineer",
      "source": "db"
    }
  ],
  "chosen_role": {
    "display_name": "Machine Learning Engineer",
    "id": null,
    "rationale": "This role aligns with primary skills like Python, PyTorch, and multiple ML frameworks essential for machine learning tasks.",
    "role_archetype": "An expert who designs and implements machine learning models and systems.",
    "slug": "machine-learning-engineer",
    "source": "llm"
  },
  "dimensions": [
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Cloud Security Scripting \u0026 DSL Languages",
        "id": 248,
        "rationale": "Proficiency in programming and domain-specific languages used to automate and script cloud security controls.",
        "slug": "cloud-security-scripting-dsl-languages",
        "source": "db"
      },
      "input_skill": "Python",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Cloud Security Engineer",
          "id": 23,
          "rationale": null,
          "role_archetype": null,
          "slug": "cloud-security-engineer",
          "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": "Python",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Backend Developer",
          "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": "Fullstack Developer",
          "id": 15,
          "rationale": null,
          "role_archetype": null,
          "slug": "full-stack-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages and Scripting",
        "id": 59,
        "rationale": "Languages used to write security automation, analysis scripts, detection logic, and remediation helpers. This is the primary implementation surface for a cybersecurity engineer across tooling and response workflows.",
        "slug": "programming-languages-and-scripting",
        "source": "db"
      },
      "input_skill": "Python",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Cybersecurity Engineer",
          "id": 5,
          "rationale": null,
          "role_archetype": null,
          "slug": "cybersecurity-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Programming Languages for Data Work",
        "id": 21,
        "rationale": "Languages used to implement data pipelines, transformations, and operational glue. This is the primary coding surface for building ingestion, enrichment, and automation logic in data engineering.",
        "slug": "programming-languages-for-data-work",
        "source": "db"
      },
      "input_skill": "Python",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Data Engineer",
          "id": 2,
          "rationale": null,
          "role_archetype": null,
          "slug": "data-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": "Python",
      "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": "Python",
      "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": "ML Frameworks and Libraries",
        "id": 40,
        "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
        "slug": "ml-frameworks-and-libraries",
        "source": "db"
      },
      "input_skill": "PyTorch",
      "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": "Model Fine-Tuning \u0026 Adaptation",
        "id": 212,
        "rationale": "Techniques and libraries for adapting pre-trained language models to specific tasks or domains.",
        "slug": "model-fine-tuning-adaptation",
        "source": "db"
      },
      "input_skill": "PyTorch",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "ML Frameworks and Libraries",
        "id": 40,
        "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
        "slug": "ml-frameworks-and-libraries",
        "source": "db"
      },
      "input_skill": "TensorFlow",
      "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": "ML Frameworks and Libraries",
        "id": 40,
        "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
        "slug": "ml-frameworks-and-libraries",
        "source": "db"
      },
      "input_skill": "JAX",
      "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": "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": "RAG",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "ML Frameworks and Libraries",
        "id": 40,
        "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
        "slug": "ml-frameworks-and-libraries",
        "source": "db"
      },
      "input_skill": "FAISS",
      "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": "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": "FAISS",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Vector Databases",
        "id": 198,
        "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
        "slug": "vector-databases",
        "source": "db"
      },
      "input_skill": "pgvector",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-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": "A/B Testing",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Systems Programming",
        "id": 166,
        "rationale": "Systems programming covers low-level software development where performance, memory safety, and direct control over resources matter. Rust fits here because it is commonly used for OS-adjacent services, infrastructure components, and other performance-sensitive systems code.",
        "slug": "d_init_02",
        "source": "db"
      },
      "input_skill": "A/B Testing",
      "llm_role": null,
      "roles_from_db": []
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "MLOps Platforms and Lifecycle",
        "id": 43,
        "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
        "slug": "mlops-platforms-and-lifecycle",
        "source": "db"
      },
      "input_skill": "MLflow",
      "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": "CI/CD Pipeline Platforms",
        "id": 150,
        "rationale": "Systems used to define, run, and maintain automated build and deployment workflows. This cluster is coherent because the role owns delivery automation end to end, including pipeline reliability and promotion logic.",
        "slug": "ci-cd-pipeline-platforms",
        "source": "db"
      },
      "input_skill": "CI/CD",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "DevOps Engineer",
          "id": 10,
          "rationale": null,
          "role_archetype": null,
          "slug": "devops-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "CI/CD for Machine Learning",
        "id": 56,
        "rationale": "Tools and platforms for automating ML model integration, testing, and deployment pipelines.",
        "slug": "ci-cd-for-machine-learning",
        "source": "db"
      },
      "input_skill": "CI/CD",
      "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": "LLM Operations and Orchestration",
        "id": 49,
        "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
        "slug": "llm-operations-and-orchestration",
        "source": "db"
      },
      "input_skill": "Pinecone",
      "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": "Vector Databases",
        "id": 198,
        "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
        "slug": "vector-databases",
        "source": "db"
      },
      "input_skill": "Pinecone",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Vector Databases",
        "id": 198,
        "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
        "slug": "vector-databases",
        "source": "db"
      },
      "input_skill": "Weaviate",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Embedding Models",
        "id": 199,
        "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
        "slug": "embedding-models",
        "source": "db"
      },
      "input_skill": "text-embedding-3",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Embedding Models",
        "id": 199,
        "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
        "slug": "embedding-models",
        "source": "db"
      },
      "input_skill": "BGE",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "LLM Operations and Orchestration",
        "id": 49,
        "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
        "slug": "llm-operations-and-orchestration",
        "source": "db"
      },
      "input_skill": "vLLM",
      "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": "LLM Serving \u0026 Deployment",
        "id": 209,
        "rationale": "Tools and frameworks for hosting and serving LLM models in production environments.",
        "slug": "llm-serving-deployment",
        "source": "db"
      },
      "input_skill": "vLLM",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "AI Engineer",
          "id": 13,
          "rationale": null,
          "role_archetype": null,
          "slug": "ai-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Model Optimization and Acceleration",
        "id": 53,
        "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
        "slug": "model-optimization-and-acceleration",
        "source": "db"
      },
      "input_skill": "Triton",
      "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": "MLOps Platforms and Lifecycle",
        "id": 43,
        "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
        "slug": "mlops-platforms-and-lifecycle",
        "source": "db"
      },
      "input_skill": "Kubeflow",
      "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": "Experiment Tracking and Evaluation",
        "id": 44,
        "rationale": "Tools and practices for recording experiments, comparing runs, and assessing model quality before release. This dimension focuses on reproducibility, metrics, artifacts, and offline evaluation workflows.",
        "slug": "experiment-tracking-and-evaluation",
        "source": "db"
      },
      "input_skill": "Weights \u0026 Biases",
      "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": "Workflow Orchestration for ML Pipelines",
        "id": 54,
        "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
        "slug": "workflow-orchestration-for-ml-pipelines",
        "source": "db"
      },
      "input_skill": "Airflow",
      "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": "Data Pipeline Orchestration",
        "id": 23,
        "rationale": "Workflow engines that schedule, coordinate, and recover batch data jobs. This cluster covers dependency management, retries, backfills, sensors, and operational control of pipeline DAGs.",
        "slug": "data-pipeline-orchestration",
        "source": "db"
      },
      "input_skill": "Prefect",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Data Engineer",
          "id": 2,
          "rationale": null,
          "role_archetype": null,
          "slug": "data-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Workflow Orchestration for ML Pipelines",
        "id": 54,
        "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
        "slug": "workflow-orchestration-for-ml-pipelines",
        "source": "db"
      },
      "input_skill": "Prefect",
      "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": "Container Orchestration Platforms",
        "id": 134,
        "rationale": "Platforms that schedule and manage containerized workloads across clusters and environments. Cloud Architects need these to define workload placement standards, cluster boundaries, and platform capabilities.",
        "slug": "container-orchestration-platforms",
        "source": "db"
      },
      "input_skill": "Kubernetes",
      "llm_role": null,
      "roles_from_db": [
        {
          "display_name": "Cloud Architect",
          "id": 9,
          "rationale": null,
          "role_archetype": null,
          "slug": "cloud-architect",
          "source": "db"
        },
        {
          "display_name": "DevOps Engineer",
          "id": 10,
          "rationale": null,
          "role_archetype": null,
          "slug": "devops-engineer",
          "source": "db"
        }
      ]
    },
    {
      "dimension": {
        "difficulty_hint": "well_known",
        "display_name": "Kubernetes for ML Workloads",
        "id": 47,
        "rationale": "Kubernetes-native components used to schedule, accelerate, and isolate ML training and serving workloads. This includes GPU enablement and ML-specific controllers rather than generic cluster administration.",
        "slug": "kubernetes-for-ml-workloads",
        "source": "db"
      },
      "input_skill": "Kubernetes",
      "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": "Model Optimization and Acceleration",
        "id": 53,
        "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
        "slug": "model-optimization-and-acceleration",
        "source": "db"
      },
      "input_skill": "Quantization",
      "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": "Video Compression Algorithms",
        "id": 226,
        "rationale": "Core compression and decompression techniques that determine bitrate, quality, and decode cost. This cluster covers the algorithmic heart of codec engineering.",
        "slug": "video-compression-algorithms",
        "source": "db"
      },
      "input_skill": "Quantization",
      "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": [
    "Python",
    "PyTorch",
    "TensorFlow",
    "JAX",
    "Hugging Face Transformers",
    "Llama 3",
    "Mistral",
    "RAG",
    "FAISS",
    "pgvector",
    "A/B Testing",
    "MLflow",
    "CI/CD",
    "Pinecone",
    "Weaviate",
    "text-embedding-3",
    "BGE",
    "vLLM",
    "TGI",
    "Triton",
    "Kubeflow",
    "Weights \u0026 Biases",
    "Airflow",
    "Prefect",
    "Kubernetes",
    "GPU scheduling",
    "BM25",
    "Quantization",
    "FP16",
    "INT8"
  ],
  "input_llm_skills": [
    "Python",
    "PyTorch",
    "TensorFlow",
    "JAX",
    "Hugging Face Transformers",
    "Llama 3",
    "Mistral",
    "RAG",
    "FAISS",
    "pgvector",
    "A/B Testing",
    "MLflow",
    "CI/CD",
    "Pinecone",
    "Weaviate",
    "text-embedding-3",
    "BGE",
    "vLLM",
    "TGI",
    "Triton",
    "Kubeflow",
    "Weights \u0026 Biases",
    "Airflow",
    "Prefect",
    "Kubernetes",
    "GPU scheduling",
    "BM25",
    "Quantization",
    "FP16",
    "INT8"
  ],
  "new_aliases_persisted": 0,
  "run_id": "0aa94386-7f59-430c-b6f7-8f25e5b5c76e",
  "skills_detail": [
    {
      "aliases_in_db": [
        {
          "alias_text": "Python",
          "alias_type": "CANONICAL",
          "id": 67,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 2",
          "alias_type": "VERSION",
          "id": 72,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 2.x",
          "alias_type": "VERSION",
          "id": 74,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 3",
          "alias_type": "VERSION",
          "id": 73,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 3.10",
          "alias_type": "VERSION",
          "id": 76,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 3.11",
          "alias_type": "VERSION",
          "id": 77,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 3.12",
          "alias_type": "VERSION",
          "id": 78,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Python 3.x",
          "alias_type": "VERSION",
          "id": 75,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "py",
          "alias_type": "VERSION",
          "id": 2183,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "py2",
          "alias_type": "VERSION",
          "id": 68,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "py3",
          "alias_type": "VERSION",
          "id": 69,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "python 3",
          "alias_type": "VERSION",
          "id": 2186,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "python 3.x",
          "alias_type": "VERSION",
          "id": 2849,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "python2",
          "alias_type": "VERSION",
          "id": 70,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "python3",
          "alias_type": "VERSION",
          "id": 71,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "python3.x",
          "alias_type": "VERSION",
          "id": 2848,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 6,
        "display_name": "Python",
        "id": 5,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "python",
        "sub_category_id": 96,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Cloud Security Scripting \u0026 DSL Languages",
            "id": 248,
            "rationale": "Proficiency in programming and domain-specific languages used to automate and script cloud security controls.",
            "slug": "cloud-security-scripting-dsl-languages",
            "source": "db"
          },
          "input_skill": "Python",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Cloud Security Engineer",
              "id": 23,
              "rationale": null,
              "role_archetype": null,
              "slug": "cloud-security-engineer",
              "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": "Python",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Backend Developer",
              "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": "Fullstack Developer",
              "id": 15,
              "rationale": null,
              "role_archetype": null,
              "slug": "full-stack-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages and Scripting",
            "id": 59,
            "rationale": "Languages used to write security automation, analysis scripts, detection logic, and remediation helpers. This is the primary implementation surface for a cybersecurity engineer across tooling and response workflows.",
            "slug": "programming-languages-and-scripting",
            "source": "db"
          },
          "input_skill": "Python",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Cybersecurity Engineer",
              "id": 5,
              "rationale": null,
              "role_archetype": null,
              "slug": "cybersecurity-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Programming Languages for Data Work",
            "id": 21,
            "rationale": "Languages used to implement data pipelines, transformations, and operational glue. This is the primary coding surface for building ingestion, enrichment, and automation logic in data engineering.",
            "slug": "programming-languages-for-data-work",
            "source": "db"
          },
          "input_skill": "Python",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Data Engineer",
              "id": 2,
              "rationale": null,
              "role_archetype": null,
              "slug": "data-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": "Python",
          "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": "Python",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AR/VR Engineer",
              "id": 8,
              "rationale": null,
              "role_archetype": null,
              "slug": "ar-vr-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Python",
      "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": "PyTorch",
          "alias_type": "CANONICAL",
          "id": 441,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "PyTorch",
        "id": 195,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "pytorch",
        "sub_category_id": 156,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "ML Frameworks and Libraries",
            "id": 40,
            "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
            "slug": "ml-frameworks-and-libraries",
            "source": "db"
          },
          "input_skill": "PyTorch",
          "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": "Model Fine-Tuning \u0026 Adaptation",
            "id": 212,
            "rationale": "Techniques and libraries for adapting pre-trained language models to specific tasks or domains.",
            "slug": "model-fine-tuning-adaptation",
            "source": "db"
          },
          "input_skill": "PyTorch",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "PyTorch",
      "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": "TensorFlow",
          "alias_type": "CANONICAL",
          "id": 442,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TF1",
          "alias_type": "VERSION",
          "id": 443,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TF2",
          "alias_type": "VERSION",
          "id": 444,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TensorFlow 1",
          "alias_type": "VERSION",
          "id": 445,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TensorFlow 1.x",
          "alias_type": "VERSION",
          "id": 447,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TensorFlow 2",
          "alias_type": "VERSION",
          "id": 446,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "TensorFlow 2.x",
          "alias_type": "VERSION",
          "id": 448,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow 1",
          "alias_type": "VERSION",
          "id": 2490,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow 1.x",
          "alias_type": "VERSION",
          "id": 2494,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow 2",
          "alias_type": "VERSION",
          "id": 2491,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow 2.x",
          "alias_type": "VERSION",
          "id": 2495,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow v1",
          "alias_type": "VERSION",
          "id": 2492,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tensorflow v2",
          "alias_type": "VERSION",
          "id": 2493,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tf",
          "alias_type": "VERSION",
          "id": 2487,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tf1",
          "alias_type": "VERSION",
          "id": 2488,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "tf2",
          "alias_type": "VERSION",
          "id": 2489,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "TensorFlow",
        "id": 196,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "tensorflow",
        "sub_category_id": 156,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "ML Frameworks and Libraries",
            "id": 40,
            "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
            "slug": "ml-frameworks-and-libraries",
            "source": "db"
          },
          "input_skill": "TensorFlow",
          "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"
            }
          ]
        }
      ],
      "input_skill": "TensorFlow",
      "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": "JAX",
          "alias_type": "CANONICAL",
          "id": 451,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "JAX",
        "id": 199,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "jax",
        "sub_category_id": 156,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "ML Frameworks and Libraries",
            "id": 40,
            "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
            "slug": "ml-frameworks-and-libraries",
            "source": "db"
          },
          "input_skill": "JAX",
          "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"
            }
          ]
        }
      ],
      "input_skill": "JAX",
      "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": [],
      "input_skill": "Hugging Face Transformers",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "TOOL",
          "sub_category": "general",
          "typical_lifespan": "MULTI_YEAR",
          "version_strategy": "UNVERSIONED",
          "volatility": "MEDIUM"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "hugging-face-transformers",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [],
      "input_skill": "Llama 3",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "TOOL",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "VERSIONED",
          "volatility": "FAST"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "llama-3",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [],
      "input_skill": "Mistral",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "TOOL",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "VERSIONED",
          "volatility": "FAST"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "mistral",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "RAG",
          "alias_type": "CANONICAL",
          "id": 1830,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "RAG",
        "id": 1194,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "rag",
        "sub_category_id": 904,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "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": "RAG",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "RAG",
      "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": "FAISS",
          "alias_type": "CANONICAL",
          "id": 1820,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "FAISS",
        "id": 1184,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "faiss",
        "sub_category_id": 894,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "ML Frameworks and Libraries",
            "id": 40,
            "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
            "slug": "ml-frameworks-and-libraries",
            "source": "db"
          },
          "input_skill": "FAISS",
          "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": "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": "FAISS",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "FAISS",
      "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": "pgvector",
          "alias_type": "CANONICAL",
          "id": 1882,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "pgvector",
        "id": 1246,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "pgvector",
        "sub_category_id": 972,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Vector Databases",
            "id": 198,
            "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
            "slug": "vector-databases",
            "source": "db"
          },
          "input_skill": "pgvector",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "pgvector",
      "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": "A/B Testing",
          "alias_type": "CANONICAL",
          "id": 2565,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 8,
        "display_name": "A/B Testing",
        "id": 1613,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "METHODOLOGY",
        "slug": "a-b-testing",
        "sub_category_id": 1214,
        "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": "A/B Testing",
          "llm_role": null,
          "roles_from_db": []
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Systems Programming",
            "id": 166,
            "rationale": "Systems programming covers low-level software development where performance, memory safety, and direct control over resources matter. Rust fits here because it is commonly used for OS-adjacent services, infrastructure components, and other performance-sensitive systems code.",
            "slug": "d_init_02",
            "source": "db"
          },
          "input_skill": "A/B Testing",
          "llm_role": null,
          "roles_from_db": []
        }
      ],
      "input_skill": "A/B Testing",
      "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": "MLflow",
          "alias_type": "CANONICAL",
          "id": 470,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 13,
        "display_name": "MLflow",
        "id": 214,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "mlflow",
        "sub_category_id": 1168,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "MLOps Platforms and Lifecycle",
            "id": 43,
            "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
            "slug": "mlops-platforms-and-lifecycle",
            "source": "db"
          },
          "input_skill": "MLflow",
          "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"
            }
          ]
        }
      ],
      "input_skill": "MLflow",
      "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": "CI/CD",
          "alias_type": "CANONICAL",
          "id": 1826,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 8,
        "display_name": "CI/CD",
        "id": 1190,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "METHODOLOGY",
        "slug": "ci-cd",
        "sub_category_id": 900,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "CI/CD Pipeline Platforms",
            "id": 150,
            "rationale": "Systems used to define, run, and maintain automated build and deployment workflows. This cluster is coherent because the role owns delivery automation end to end, including pipeline reliability and promotion logic.",
            "slug": "ci-cd-pipeline-platforms",
            "source": "db"
          },
          "input_skill": "CI/CD",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "DevOps Engineer",
              "id": 10,
              "rationale": null,
              "role_archetype": null,
              "slug": "devops-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "CI/CD for Machine Learning",
            "id": 56,
            "rationale": "Tools and platforms for automating ML model integration, testing, and deployment pipelines.",
            "slug": "ci-cd-for-machine-learning",
            "source": "db"
          },
          "input_skill": "CI/CD",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "CI/CD",
      "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": "Pinecone",
          "alias_type": "CANONICAL",
          "id": 503,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 9,
        "display_name": "Pinecone",
        "id": 242,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "pinecone",
        "sub_category_id": 177,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "LLM Operations and Orchestration",
            "id": 49,
            "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
            "slug": "llm-operations-and-orchestration",
            "source": "db"
          },
          "input_skill": "Pinecone",
          "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": "Vector Databases",
            "id": 198,
            "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
            "slug": "vector-databases",
            "source": "db"
          },
          "input_skill": "Pinecone",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Pinecone",
      "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": "Weaviate",
          "alias_type": "CANONICAL",
          "id": 1878,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 9,
        "display_name": "Weaviate",
        "id": 1242,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "weaviate",
        "sub_category_id": 177,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Vector Databases",
            "id": 198,
            "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
            "slug": "vector-databases",
            "source": "db"
          },
          "input_skill": "Weaviate",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Weaviate",
      "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": "text-embedding-3-large",
          "alias_type": "CANONICAL",
          "id": 1888,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 11,
        "display_name": "text-embedding-3-large",
        "id": 1252,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CLOUD_SERVICE",
        "slug": "text-embedding-3-large",
        "sub_category_id": 1005,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Embedding Models",
            "id": 199,
            "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
            "slug": "embedding-models",
            "source": "db"
          },
          "input_skill": "text-embedding-3",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "text-embedding-3",
      "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": "BGE",
          "alias_type": "CANONICAL",
          "id": 1885,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 7,
        "display_name": "BGE",
        "id": 1249,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LIBRARY",
        "slug": "bge",
        "sub_category_id": 973,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Embedding Models",
            "id": 199,
            "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
            "slug": "embedding-models",
            "source": "db"
          },
          "input_skill": "BGE",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "BGE",
      "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": "vLLM",
          "alias_type": "CANONICAL",
          "id": 502,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 13,
        "display_name": "vLLM",
        "id": 241,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "vllm",
        "sub_category_id": 1167,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "LLM Operations and Orchestration",
            "id": 49,
            "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
            "slug": "llm-operations-and-orchestration",
            "source": "db"
          },
          "input_skill": "vLLM",
          "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": "LLM Serving \u0026 Deployment",
            "id": 209,
            "rationale": "Tools and frameworks for hosting and serving LLM models in production environments.",
            "slug": "llm-serving-deployment",
            "source": "db"
          },
          "input_skill": "vLLM",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "AI Engineer",
              "id": 13,
              "rationale": null,
              "role_archetype": null,
              "slug": "ai-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "vLLM",
      "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": [],
      "input_skill": "TGI",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "TOOL",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "VERSIONED",
          "volatility": "FAST"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "tgi",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "Triton",
          "alias_type": "CANONICAL",
          "id": 2026,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 6,
        "display_name": "Triton",
        "id": 1367,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "LANGUAGE",
        "slug": "triton",
        "sub_category_id": 1033,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Model Optimization and Acceleration",
            "id": 53,
            "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
            "slug": "model-optimization-and-acceleration",
            "source": "db"
          },
          "input_skill": "Triton",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "ML Engineer",
              "id": 3,
              "rationale": null,
              "role_archetype": null,
              "slug": "ml-engineer",
              "source": "db"
            }
          ]
        }
      ],
      "input_skill": "Triton",
      "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": "Kubeflow",
          "alias_type": "CANONICAL",
          "id": 465,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubeflow 1.x",
          "alias_type": "VERSION",
          "id": 468,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubeflow 2.x",
          "alias_type": "VERSION",
          "id": 469,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubeflow v1",
          "alias_type": "VERSION",
          "id": 466,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubeflow v2",
          "alias_type": "VERSION",
          "id": 467,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 5,
        "display_name": "Kubeflow",
        "id": 213,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "FRAMEWORK",
        "slug": "kubeflow",
        "sub_category_id": 1127,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "MLOps Platforms and Lifecycle",
            "id": 43,
            "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
            "slug": "mlops-platforms-and-lifecycle",
            "source": "db"
          },
          "input_skill": "Kubeflow",
          "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"
            }
          ]
        }
      ],
      "input_skill": "Kubeflow",
      "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": "Weights \u0026 Biases",
          "alias_type": "CANONICAL",
          "id": 471,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 9,
        "display_name": "Weights \u0026 Biases",
        "id": 215,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "weights-biases",
        "sub_category_id": 171,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Experiment Tracking and Evaluation",
            "id": 44,
            "rationale": "Tools and practices for recording experiments, comparing runs, and assessing model quality before release. This dimension focuses on reproducibility, metrics, artifacts, and offline evaluation workflows.",
            "slug": "experiment-tracking-and-evaluation",
            "source": "db"
          },
          "input_skill": "Weights \u0026 Biases",
          "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"
            }
          ]
        }
      ],
      "input_skill": "Weights \u0026 Biases",
      "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": "Airflow",
          "alias_type": "CANONICAL",
          "id": 526,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "airflow 2",
          "alias_type": "VERSION",
          "id": 2477,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "airflow-2",
          "alias_type": "VERSION",
          "id": 2478,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "airflow2",
          "alias_type": "VERSION",
          "id": 2476,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "airflow2.x",
          "alias_type": "VERSION",
          "id": 2479,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "apache airflow 2",
          "alias_type": "VERSION",
          "id": 2480,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 13,
        "display_name": "Airflow",
        "id": 265,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "airflow",
        "sub_category_id": 130,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Workflow Orchestration for ML Pipelines",
            "id": 54,
            "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
            "slug": "workflow-orchestration-for-ml-pipelines",
            "source": "db"
          },
          "input_skill": "Airflow",
          "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"
            }
          ]
        }
      ],
      "input_skill": "Airflow",
      "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": "Prefect",
          "alias_type": "CANONICAL",
          "id": 306,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 13,
        "display_name": "Prefect",
        "id": 112,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "TOOL",
        "slug": "prefect",
        "sub_category_id": 130,
        "typical_lifespan": "EVERGREEN",
        "volatility": "EMERGING"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Data Pipeline Orchestration",
            "id": 23,
            "rationale": "Workflow engines that schedule, coordinate, and recover batch data jobs. This cluster covers dependency management, retries, backfills, sensors, and operational control of pipeline DAGs.",
            "slug": "data-pipeline-orchestration",
            "source": "db"
          },
          "input_skill": "Prefect",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Data Engineer",
              "id": 2,
              "rationale": null,
              "role_archetype": null,
              "slug": "data-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Workflow Orchestration for ML Pipelines",
            "id": 54,
            "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
            "slug": "workflow-orchestration-for-ml-pipelines",
            "source": "db"
          },
          "input_skill": "Prefect",
          "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"
            }
          ]
        }
      ],
      "input_skill": "Prefect",
      "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": "Kubernetes",
          "alias_type": "CANONICAL",
          "id": 1267,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubernetes 1.0+",
          "alias_type": "VERSION",
          "id": 1271,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubernetes 1.x",
          "alias_type": "VERSION",
          "id": 1270,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "Kubernetes v1",
          "alias_type": "VERSION",
          "id": 1269,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "k8s",
          "alias_type": "VERSION",
          "id": 1268,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "kubernetes 1.x",
          "alias_type": "VERSION",
          "id": 1400,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        },
        {
          "alias_text": "kubernetes latest",
          "alias_type": "VERSION",
          "id": 1401,
          "is_primary": false,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 9,
        "display_name": "Kubernetes",
        "id": 726,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "PLATFORM",
        "slug": "kubernetes",
        "sub_category_id": 557,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Container Orchestration Platforms",
            "id": 134,
            "rationale": "Platforms that schedule and manage containerized workloads across clusters and environments. Cloud Architects need these to define workload placement standards, cluster boundaries, and platform capabilities.",
            "slug": "container-orchestration-platforms",
            "source": "db"
          },
          "input_skill": "Kubernetes",
          "llm_role": null,
          "roles_from_db": [
            {
              "display_name": "Cloud Architect",
              "id": 9,
              "rationale": null,
              "role_archetype": null,
              "slug": "cloud-architect",
              "source": "db"
            },
            {
              "display_name": "DevOps Engineer",
              "id": 10,
              "rationale": null,
              "role_archetype": null,
              "slug": "devops-engineer",
              "source": "db"
            }
          ]
        },
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Kubernetes for ML Workloads",
            "id": 47,
            "rationale": "Kubernetes-native components used to schedule, accelerate, and isolate ML training and serving workloads. This includes GPU enablement and ML-specific controllers rather than generic cluster administration.",
            "slug": "kubernetes-for-ml-workloads",
            "source": "db"
          },
          "input_skill": "Kubernetes",
          "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"
            }
          ]
        }
      ],
      "input_skill": "Kubernetes",
      "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": [],
      "input_skill": "GPU scheduling",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Infrastructure Tools",
          "skill_nature": "CONCEPT",
          "sub_category": "general",
          "typical_lifespan": "MULTI_YEAR",
          "version_strategy": "UNVERSIONED",
          "volatility": "MEDIUM"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "gpu-scheduling",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [],
      "input_skill": "BM25",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Information Retrieval",
          "skill_nature": "CONCEPT",
          "sub_category": "general",
          "typical_lifespan": "EVERGREEN",
          "version_strategy": "UNVERSIONED",
          "volatility": "STABLE"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "bm25",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [
        {
          "alias_text": "quantization",
          "alias_type": "CANONICAL",
          "id": 521,
          "is_primary": true,
          "match_strategy": "CASE_INSENSITIVE"
        }
      ],
      "canonical": {
        "category_id": 2,
        "display_name": "quantization",
        "id": 260,
        "is_also_category": false,
        "is_extractable": true,
        "skill_nature": "CONCEPT",
        "slug": "quantization",
        "sub_category_id": 1282,
        "typical_lifespan": "EVERGREEN",
        "volatility": "STABLE"
      },
      "dimensions": [
        {
          "dimension": {
            "difficulty_hint": "well_known",
            "display_name": "Model Optimization and Acceleration",
            "id": 53,
            "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
            "slug": "model-optimization-and-acceleration",
            "source": "db"
          },
          "input_skill": "Quantization",
          "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": "Video Compression Algorithms",
            "id": 226,
            "rationale": "Core compression and decompression techniques that determine bitrate, quality, and decode cost. This cluster covers the algorithmic heart of codec engineering.",
            "slug": "video-compression-algorithms",
            "source": "db"
          },
          "input_skill": "Quantization",
          "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": "Quantization",
      "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": [],
      "input_skill": "FP16",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "CONCEPT",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "VERSIONED",
          "volatility": "FAST"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "fp16",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    },
    {
      "aliases_in_db": [],
      "canonical": null,
      "dimensions": [],
      "input_skill": "INT8",
      "matched_via": null,
      "new_alias_persisted": false,
      "new_alias_text": null,
      "new_skill_meta": {
        "derived": {
          "category": "Machine Learning Frameworks",
          "skill_nature": "CONCEPT",
          "sub_category": "general",
          "typical_lifespan": "SHORT_LIVED",
          "version_strategy": "VERSIONED",
          "volatility": "FAST"
        },
        "enrichment": null,
        "keep_log": [],
        "locked_dimensions": [],
        "merge_log": [],
        "placed": null,
        "relationships": null,
        "skill_id": "int8",
        "split_log": [],
        "typed": null,
        "warnings": []
      },
      "source_tag": "llm",
      "was_in_llm_skills": true
    }
  ],
  "unmatched_skills": [
    "Hugging Face Transformers",
    "Llama 3",
    "Mistral",
    "TGI",
    "GPU scheduling",
    "BM25",
    "FP16",
    "INT8"
  ]
}

{
  "chosen_role": {
    "display_name": "ML Engineer",
    "id": 3,
    "rationale": "This role aligns with primary skills like Python, PyTorch, and multiple ML frameworks essential for machine learning tasks.",
    "role_archetype": "An expert who designs and implements machine learning models and systems.",
    "slug": "ml-engineer",
    "source": "db"
  },
  "chosen_role_resolution": "in_db",
  "final_input_skills": [
    {
      "skill": "Python",
      "tag": "in_db"
    },
    {
      "skill": "PyTorch",
      "tag": "in_db"
    },
    {
      "skill": "TensorFlow",
      "tag": "in_db"
    },
    {
      "skill": "JAX",
      "tag": "in_db"
    },
    {
      "skill": "Hugging Face Transformers",
      "tag": "new"
    },
    {
      "skill": "Llama 3",
      "tag": "new"
    },
    {
      "skill": "Mistral",
      "tag": "new"
    },
    {
      "skill": "RAG",
      "tag": "in_db"
    },
    {
      "skill": "FAISS",
      "tag": "in_db"
    },
    {
      "skill": "pgvector",
      "tag": "in_db"
    },
    {
      "skill": "A/B Testing",
      "tag": "in_db"
    },
    {
      "skill": "MLflow",
      "tag": "in_db"
    },
    {
      "skill": "CI/CD",
      "tag": "in_db"
    },
    {
      "skill": "Pinecone",
      "tag": "in_db"
    },
    {
      "skill": "Weaviate",
      "tag": "in_db"
    },
    {
      "skill": "text-embedding-3",
      "tag": "in_db"
    },
    {
      "skill": "BGE",
      "tag": "in_db"
    },
    {
      "skill": "vLLM",
      "tag": "in_db"
    },
    {
      "skill": "TGI",
      "tag": "new"
    },
    {
      "skill": "Triton",
      "tag": "in_db"
    },
    {
      "skill": "Kubeflow",
      "tag": "in_db"
    },
    {
      "skill": "Weights \u0026 Biases",
      "tag": "in_db"
    },
    {
      "skill": "Airflow",
      "tag": "in_db"
    },
    {
      "skill": "Prefect",
      "tag": "in_db"
    },
    {
      "skill": "Kubernetes",
      "tag": "in_db"
    },
    {
      "skill": "GPU scheduling",
      "tag": "new"
    },
    {
      "skill": "BM25",
      "tag": "new"
    },
    {
      "skill": "Quantization",
      "tag": "in_db"
    },
    {
      "skill": "FP16",
      "tag": "new"
    },
    {
      "skill": "INT8",
      "tag": "new"
    }
  ],
  "llm_cost_api1_usd": null,
  "llm_cost_api2_usd": null,
  "llm_cost_api3_usd": null,
  "llm_cost_total_usd": null,
  "persistence": {
    "items": [
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Cloud Security Scripting \u0026 DSL Languages",
          "id": 248,
          "rationale": "Proficiency in programming and domain-specific languages used to automate and script cloud security controls.",
          "slug": "cloud-security-scripting-dsl-languages",
          "source": "db"
        },
        "dimension_id": 248,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Cloud Security Engineer",
            "id": 23,
            "rationale": null,
            "role_archetype": null,
            "slug": "cloud-security-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 1,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Backend Developer",
            "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": "Fullstack Developer",
            "id": 15,
            "rationale": null,
            "role_archetype": null,
            "slug": "full-stack-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Programming Languages and Scripting",
          "id": 59,
          "rationale": "Languages used to write security automation, analysis scripts, detection logic, and remediation helpers. This is the primary implementation surface for a cybersecurity engineer across tooling and response workflows.",
          "slug": "programming-languages-and-scripting",
          "source": "db"
        },
        "dimension_id": 59,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Cybersecurity Engineer",
            "id": 5,
            "rationale": null,
            "role_archetype": null,
            "slug": "cybersecurity-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Programming Languages for Data Work",
          "id": 21,
          "rationale": "Languages used to implement data pipelines, transformations, and operational glue. This is the primary coding surface for building ingestion, enrichment, and automation logic in data engineering.",
          "slug": "programming-languages-for-data-work",
          "source": "db"
        },
        "dimension_id": 21,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Data Engineer",
            "id": 2,
            "rationale": null,
            "role_archetype": null,
            "slug": "data-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 39,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 97,
        "input_skill": "Python",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AR/VR Engineer",
            "id": 8,
            "rationale": null,
            "role_archetype": null,
            "slug": "ar-vr-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 5,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "ML Frameworks and Libraries",
          "id": 40,
          "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
          "slug": "ml-frameworks-and-libraries",
          "source": "db"
        },
        "dimension_id": 40,
        "input_skill": "PyTorch",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 195,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Model Fine-Tuning \u0026 Adaptation",
          "id": 212,
          "rationale": "Techniques and libraries for adapting pre-trained language models to specific tasks or domains.",
          "slug": "model-fine-tuning-adaptation",
          "source": "db"
        },
        "dimension_id": 212,
        "input_skill": "PyTorch",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 195,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "ML Frameworks and Libraries",
          "id": 40,
          "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
          "slug": "ml-frameworks-and-libraries",
          "source": "db"
        },
        "dimension_id": 40,
        "input_skill": "TensorFlow",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 196,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "ML Frameworks and Libraries",
          "id": 40,
          "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
          "slug": "ml-frameworks-and-libraries",
          "source": "db"
        },
        "dimension_id": 40,
        "input_skill": "JAX",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 199,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 96,
        "input_skill": "RAG",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [],
        "skill_dimension_saved": true,
        "skill_id": 1194,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "ML Frameworks and Libraries",
          "id": 40,
          "rationale": "Core libraries used to define models, train them, run inference, and evaluate predictive performance. These frameworks shape how ML engineers express model architectures and training loops.",
          "slug": "ml-frameworks-and-libraries",
          "source": "db"
        },
        "dimension_id": 40,
        "input_skill": "FAISS",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1184,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 96,
        "input_skill": "FAISS",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [],
        "skill_dimension_saved": true,
        "skill_id": 1184,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Vector Databases",
          "id": 198,
          "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
          "slug": "vector-databases",
          "source": "db"
        },
        "dimension_id": 198,
        "input_skill": "pgvector",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1246,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "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"
        },
        "dimension_id": 96,
        "input_skill": "A/B Testing",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [],
        "skill_dimension_saved": true,
        "skill_id": 1613,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Systems Programming",
          "id": 166,
          "rationale": "Systems programming covers low-level software development where performance, memory safety, and direct control over resources matter. Rust fits here because it is commonly used for OS-adjacent services, infrastructure components, and other performance-sensitive systems code.",
          "slug": "d_init_02",
          "source": "db"
        },
        "dimension_id": 166,
        "input_skill": "A/B Testing",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [],
        "skill_dimension_saved": true,
        "skill_id": 1613,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "MLOps Platforms and Lifecycle",
          "id": 43,
          "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
          "slug": "mlops-platforms-and-lifecycle",
          "source": "db"
        },
        "dimension_id": 43,
        "input_skill": "MLflow",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 214,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "CI/CD Pipeline Platforms",
          "id": 150,
          "rationale": "Systems used to define, run, and maintain automated build and deployment workflows. This cluster is coherent because the role owns delivery automation end to end, including pipeline reliability and promotion logic.",
          "slug": "ci-cd-pipeline-platforms",
          "source": "db"
        },
        "dimension_id": 150,
        "input_skill": "CI/CD",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "DevOps Engineer",
            "id": 10,
            "rationale": null,
            "role_archetype": null,
            "slug": "devops-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1190,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "CI/CD for Machine Learning",
          "id": 56,
          "rationale": "Tools and platforms for automating ML model integration, testing, and deployment pipelines.",
          "slug": "ci-cd-for-machine-learning",
          "source": "db"
        },
        "dimension_id": 56,
        "input_skill": "CI/CD",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "ML Engineer",
            "id": 3,
            "rationale": null,
            "role_archetype": null,
            "slug": "ml-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1190,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "LLM Operations and Orchestration",
          "id": 49,
          "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
          "slug": "llm-operations-and-orchestration",
          "source": "db"
        },
        "dimension_id": 49,
        "input_skill": "Pinecone",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 242,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Vector Databases",
          "id": 198,
          "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
          "slug": "vector-databases",
          "source": "db"
        },
        "dimension_id": 198,
        "input_skill": "Pinecone",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 242,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Vector Databases",
          "id": 198,
          "rationale": "Specialized storage and indexing systems used to persist embeddings and support similarity search. This is a distinct vendor-family cluster because AI features often depend on a concrete vector store choice and its operational behavior.",
          "slug": "vector-databases",
          "source": "db"
        },
        "dimension_id": 198,
        "input_skill": "Weaviate",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1242,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Embedding Models",
          "id": 199,
          "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
          "slug": "embedding-models",
          "source": "db"
        },
        "dimension_id": 199,
        "input_skill": "text-embedding-3",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Skipped \u2014 no persistable v3 meta for new skill",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": false,
        "skill_id": null,
        "skill_tag": "new",
        "skipped_reason": "skill_not_in_db_v3_proposed"
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Embedding Models",
          "id": 199,
          "rationale": "Models and APIs used to convert text or other content into vector representations for retrieval, clustering, and semantic matching. This cluster is coherent because embedding choice affects retrieval quality, cost, and latency in AI features.",
          "slug": "embedding-models",
          "source": "db"
        },
        "dimension_id": 199,
        "input_skill": "BGE",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1249,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "LLM Operations and Orchestration",
          "id": 49,
          "rationale": "Operational stack for building, serving, evaluating, and orchestrating LLM-based systems. This includes vector retrieval, prompt workflows, LLM serving, and observability for generative applications.",
          "slug": "llm-operations-and-orchestration",
          "source": "db"
        },
        "dimension_id": 49,
        "input_skill": "vLLM",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 241,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "LLM Serving \u0026 Deployment",
          "id": 209,
          "rationale": "Tools and frameworks for hosting and serving LLM models in production environments.",
          "slug": "llm-serving-deployment",
          "source": "db"
        },
        "dimension_id": 209,
        "input_skill": "vLLM",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "AI Engineer",
            "id": 13,
            "rationale": null,
            "role_archetype": null,
            "slug": "ai-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 241,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Model Optimization and Acceleration",
          "id": 53,
          "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
          "slug": "model-optimization-and-acceleration",
          "source": "db"
        },
        "dimension_id": 53,
        "input_skill": "Triton",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "ML Engineer",
            "id": 3,
            "rationale": null,
            "role_archetype": null,
            "slug": "ml-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 1367,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "MLOps Platforms and Lifecycle",
          "id": 43,
          "rationale": "End-to-end managed platforms used to train, deploy, register, and govern models across their lifecycle. This is the operational control plane for production ML workflows.",
          "slug": "mlops-platforms-and-lifecycle",
          "source": "db"
        },
        "dimension_id": 43,
        "input_skill": "Kubeflow",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 213,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Experiment Tracking and Evaluation",
          "id": 44,
          "rationale": "Tools and practices for recording experiments, comparing runs, and assessing model quality before release. This dimension focuses on reproducibility, metrics, artifacts, and offline evaluation workflows.",
          "slug": "experiment-tracking-and-evaluation",
          "source": "db"
        },
        "dimension_id": 44,
        "input_skill": "Weights \u0026 Biases",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 215,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Workflow Orchestration for ML Pipelines",
          "id": 54,
          "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
          "slug": "workflow-orchestration-for-ml-pipelines",
          "source": "db"
        },
        "dimension_id": 54,
        "input_skill": "Airflow",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 265,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Data Pipeline Orchestration",
          "id": 23,
          "rationale": "Workflow engines that schedule, coordinate, and recover batch data jobs. This cluster covers dependency management, retries, backfills, sensors, and operational control of pipeline DAGs.",
          "slug": "data-pipeline-orchestration",
          "source": "db"
        },
        "dimension_id": 23,
        "input_skill": "Prefect",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Data Engineer",
            "id": 2,
            "rationale": null,
            "role_archetype": null,
            "slug": "data-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 112,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Workflow Orchestration for ML Pipelines",
          "id": 54,
          "rationale": "Workflow engines used to coordinate training, evaluation, deployment, and retraining jobs. This cluster covers dependencies, retries, scheduling, and pipeline composition for ML lifecycle automation.",
          "slug": "workflow-orchestration-for-ml-pipelines",
          "source": "db"
        },
        "dimension_id": 54,
        "input_skill": "Prefect",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 112,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Container Orchestration Platforms",
          "id": 134,
          "rationale": "Platforms that schedule and manage containerized workloads across clusters and environments. Cloud Architects need these to define workload placement standards, cluster boundaries, and platform capabilities.",
          "slug": "container-orchestration-platforms",
          "source": "db"
        },
        "dimension_id": 134,
        "input_skill": "Kubernetes",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Cloud Architect",
            "id": 9,
            "rationale": null,
            "role_archetype": null,
            "slug": "cloud-architect",
            "source": "db"
          },
          {
            "display_name": "DevOps Engineer",
            "id": 10,
            "rationale": null,
            "role_archetype": null,
            "slug": "devops-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 726,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Kubernetes for ML Workloads",
          "id": 47,
          "rationale": "Kubernetes-native components used to schedule, accelerate, and isolate ML training and serving workloads. This includes GPU enablement and ML-specific controllers rather than generic cluster administration.",
          "slug": "kubernetes-for-ml-workloads",
          "source": "db"
        },
        "dimension_id": 47,
        "input_skill": "Kubernetes",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "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"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 726,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Model Optimization and Acceleration",
          "id": 53,
          "rationale": "Techniques for improving inference latency, throughput, memory use, and training efficiency. ML engineers use these methods to meet production constraints without sacrificing too much quality.",
          "slug": "model-optimization-and-acceleration",
          "source": "db"
        },
        "dimension_id": 53,
        "input_skill": "Quantization",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "ML Engineer",
            "id": 3,
            "rationale": null,
            "role_archetype": null,
            "slug": "ml-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 260,
        "skill_tag": "in_db",
        "skipped_reason": null
      },
      {
        "chosen_role_id": 3,
        "dimension": {
          "difficulty_hint": "well_known",
          "display_name": "Video Compression Algorithms",
          "id": 226,
          "rationale": "Core compression and decompression techniques that determine bitrate, quality, and decode cost. This cluster covers the algorithmic heart of codec engineering.",
          "slug": "video-compression-algorithms",
          "source": "db"
        },
        "dimension_id": 226,
        "input_skill": "Quantization",
        "llm_role": null,
        "matched_chosen_role": false,
        "outcome_line": "Existing dimension (library) \u00b7 Role\u2194dimension skipped (dimension not under chosen role)",
        "role_dimension_saved": false,
        "roles_from_db": [
          {
            "display_name": "Video Codec Engineer",
            "id": 22,
            "rationale": null,
            "role_archetype": null,
            "slug": "video-codec-engineer",
            "source": "db"
          }
        ],
        "skill_dimension_saved": true,
        "skill_id": 260,
        "skill_tag": "in_db",
        "skipped_reason": null
      }
    ],
    "new_skills_created": 0,
    "role_dimension_saved": 0,
    "skill_dimension_saved": 0,
    "skipped": 1
  },
  "planner_output": null,
  "run_id": "0aa94386-7f59-430c-b6f7-8f25e5b5c76e"
}