Senior MLOps Architect

Imported from vivek-doshi/mlops-playbook (.ai/skills/senior-mlops-architect/SKILL.md). Install upstream with npx skills add vivek-doshi/mlops-playbook --skill senior-mlops-architect. Copyright stays with the author.

Purpose

Use this skill for MLOps architecture decisions, lifecycle standards, and cross-repo integration with the platform foundation.

Use This Skill When

• Designing experiment tracking, model registry, and serving topology.
• Choosing defaults for data versioning, model serving, and drift monitoring.
• Defining CI/CD pathways for train, evaluate, approve, and deploy.
• Enforcing governance boundaries between platform and MLOps layers.

Repository Context To Read First

1. .ai/context/repo-summary.md
2. .ai/context/project_details.md
3. README.md
4. docs/golden-paths/
5. policy/

MLOps Architecture Rules

• Keep platform provisioning in the platform repository.
• Keep ML lifecycle tooling in this repository.
• Require promotion gates before production model serving.
• Ensure experiment and dataset lineage are reproducible.
• Treat monitoring and drift response as first-class operational flows.

Expected Outputs

• Recommended architecture path with rationale.
• Exact files to create or modify.
• Governance and risk notes.
• Validation checklist before rollout.