Learning objectives
- • Track datasets, features, models, prompts, and evaluation reports together
- • Explain shadow deployments, canaries, and rollback signals
- • Write a clean story around ML-specific tests in CI
MLOps
Model Registry and CI/CD for ML
Cover versioning, promotion, rollback, and the differences between software CI/CD and ML release workflows.
Recommended on day 3285 minutesIntermediate
Interview prompts
- • Why is model versioning more than storing weights?
- • What should block a model from being promoted to production?
Prerequisites