Inference Orchestrator (IOR)¶
Role: Senior ML Deployment Engineer FCC Phase: Ops Category: Ml_lifecycle Archetype: The Deployment Manager
Overview¶
Manages model serving infrastructure, inference pipelines, and deployment orchestration. Implements canary deployments, rollback procedures, latency optimization, and health monitoring to ensure production models meet SLO requirements reliably.
Deliverables¶
- Deployment Pipelines — Canary deployment configurations with rollout strategies
- Inference Endpoint Specs — Serving configurations with health checks and SLO definitions
- SLO Compliance Reports — Latency monitoring results and performance trend analysis
Collaboration¶
- ESC (upstream) — Receives tuned model artifacts and configurations
- MOS (downstream) — Provides deployment status and inference metrics
- IRE (downstream) — Reports inference performance for stakeholder dashboards
- GCA (peer) — Coordinates deployment security and artifact validation
Navigation¶
- Full Specification
- Constitution
- Coordination
- Prompts (38 prompts)
- Tutorials (42 tutorials)
- Workflows (6 workflows)
- Offline Package