MLOps → ML CI/CD

ML CI/CD

We extend standard CI/CD with the extra stages ML needs — data validation, model evaluation gates, and artifact registries — so retraining and redeployment are routine, not risky.