iiRecord
Agentic AI Atlas · AI Inference Cost Review
workflow:ai-inference-cost-reviewa5c.ai
II.
Workflow JSON

workflow:ai-inference-cost-review

Structured · live

AI Inference Cost Review json

Inspect the normalized record payload exactly as the atlas UI reads it.

File · workflows/workflows/workflows-finops-deep.yamlCluster · workflows
Record JSON
{
  "id": "workflow:ai-inference-cost-review",
  "_kind": "Workflow",
  "_file": "workflows/workflows/workflows-finops-deep.yaml",
  "_cluster": "workflows",
  "attributes": {
    "displayName": "AI Inference Cost Review",
    "workflowKind": "governance",
    "triggerType": "scheduled",
    "typicalCadence": "bi-weekly",
    "complexity": "cross-team",
    "description": "Reviews AI and LLM inference costs across the organization to optimize\nspend while maintaining quality -- analyzing API cost breakdowns by model,\nfeature, and team with token-level granularity, evaluating prompt\nengineering efficiency by measuring token counts against output quality\nmetrics, reviewing caching layer effectiveness including semantic cache hit\nrates and cost avoidance, assessing model selection appropriateness by\ncomparing quality-to-cost ratios across model tiers for each use case,\nidentifying opportunities to shift workloads from expensive frontier models\nto fine-tuned smaller models, tracking cost trends against usage growth to\ndetect non-linear cost scaling, reviewing batch vs real-time inference\nallocation for latency-tolerant workloads, and benchmarking per-request\ncosts against industry norms. Produces AI cost dashboard, optimization\nrecommendation report, and model-tier allocation review. Excludes model\ntraining and fine-tuning.\n"
  },
  "outgoingEdges": [
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "role:ai-champion",
      "kind": "involves_role",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "role:cloud-architect",
      "kind": "involves_role",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "role:data-scientist",
      "kind": "involves_role",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "skill-area:prompt-engineering",
      "kind": "requires_skill_area",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "skill-area:context-management",
      "kind": "requires_skill_area",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "domain:finops",
      "kind": "applies_to_domain",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "domain:operations",
      "kind": "applies_to_domain",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "responsibility:cost-optimization",
      "kind": "triggers_responsibility",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "responsibility:ai-agent-usage-review",
      "kind": "triggers_responsibility",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "org-unit:finops-team",
      "kind": "performed_by_org_unit",
      "attributes": {}
    },
    {
      "from": "workflow:ai-inference-cost-review",
      "to": "org-unit:engineering",
      "kind": "performed_by_org_unit",
      "attributes": {}
    }
  ],
  "incomingEdges": []
}