ML Model Versioning Governance overview

Governs ML model versioning, lineage tracking, and promotion workflows -- auditing model-registry completeness for metadata including training-data snapshots, hyperparameters, and evaluation metrics, validating model-card documentation against organizational standards, reviewing model-promotion gates from experimental through staging to production, enforcing reproducibility requirements by verifying training-pipeline determinism, assessing model-drift monitoring alerts and retraining trigger thresholds, auditing access controls on model artifacts and serving endpoints, and reconciling deployed model versions against the approved model inventory. Produces model governance compliance report and registry hygiene scorecard. Excludes model architecture research.