{
  "id": "skill-area:inference-optimization",
  "_kind": "SkillArea",
  "_file": "domain/skill-areas/skill-areas-expanded.yaml",
  "_cluster": "domain",
  "attributes": {
    "displayName": "Inference Optimization",
    "description": "Techniques for reducing LLM and ML inference latency and cost —\nquantization, speculative decoding, KV-cache optimization,\nbatching strategies, and hardware-aware serving tuning.\n",
    "expertiseLevels": [
      "intermediate",
      "expert"
    ]
  },
  "outgoingEdges": [
    {
      "from": "skill-area:inference-optimization",
      "to": "domain:ml-ai",
      "kind": "applies_to",
      "attributes": {
        "confidence": "primary"
      }
    },
    {
      "from": "skill-area:inference-optimization",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "primary"
      }
    },
    {
      "from": "skill-area:inference-optimization",
      "to": "skill-area:model-compression",
      "kind": "prerequisite_for_learning",
      "attributes": {
        "strength": "helpful"
      }
    }
  ],
  "incomingEdges": [
    {
      "from": "skill-area:model-compression",
      "to": "skill-area:inference-optimization",
      "kind": "prerequisite_for_learning",
      "attributes": {
        "strength": "recommended"
      }
    },
    {
      "from": "responsibility:inference-latency-sla",
      "to": "skill-area:inference-optimization",
      "kind": "requires_expertise",
      "attributes": {}
    }
  ]
}