- MLflow + Postgres + MinIO running
- Sample PyTorch/HF run logged
- Artifact persistence verified
- Standard training templates for quant and LLM teams
- Evaluation gates and promotion policies
- Model card generation and governance metadata
- Scheduled training/inference evaluation jobs
- Drift and regression alerting
- Cost/performance dashboards per model family