{
  "id": "specialization:ml-inference-serving",
  "_kind": "Specialization",
  "_file": "domain/specializations/ml-inference-serving.yaml",
  "_cluster": "domain",
  "attributes": {
    "displayName": "ML Inference Serving",
    "parentDomainId": "domain:ml-ops",
    "description": "Serving ML / LLM inference at scale — vLLM, TGI, TensorRT-LLM,\nTriton, batching, KV-cache management, and GPU autoscaling.\n"
  },
  "outgoingEdges": [
    {
      "from": "specialization:ml-inference-serving",
      "to": "domain:ml-ops",
      "kind": "specializes"
    },
    {
      "from": "specialization:ml-inference-serving",
      "to": "tool:vllm",
      "kind": "uses_tool"
    },
    {
      "from": "specialization:ml-inference-serving",
      "to": "tool:triton-inference",
      "kind": "uses_tool"
    },
    {
      "from": "specialization:ml-inference-serving",
      "to": "tool:onnx-runtime",
      "kind": "uses_tool"
    },
    {
      "from": "specialization:ml-inference-serving",
      "to": "tool:bentoml",
      "kind": "uses_tool"
    }
  ],
  "incomingEdges": [
    {
      "from": "domain:machine-learning",
      "to": "specialization:ml-inference-serving",
      "kind": "contains"
    },
    {
      "from": "domain:ml-ops",
      "to": "specialization:ml-inference-serving",
      "kind": "contains"
    },
    {
      "from": "domain:ml-ai",
      "to": "specialization:ml-inference-serving",
      "kind": "contains",
      "attributes": {}
    },
    {
      "from": "skill-area:model-serving-deployment",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "secondary"
      }
    },
    {
      "from": "skill-area:model-serving-operations",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "primary"
      }
    },
    {
      "from": "skill-area:model-robustness-testing",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "secondary"
      }
    },
    {
      "from": "skill-area:inference-performance-testing",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "primary"
      }
    },
    {
      "from": "skill-area:model-monitoring-drift-detection",
      "to": "specialization:ml-inference-serving",
      "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:model-compression",
      "to": "specialization:ml-inference-serving",
      "kind": "applies_to",
      "attributes": {
        "confidence": "secondary"
      }
    },
    {
      "from": "lib-skill:data-science-ml--inference-performance-testing",
      "to": "specialization:ml-inference-serving",
      "kind": "lib_belongs_to_specialization",
      "attributes": {
        "weight": 0.7
      }
    }
  ]
}