Whenever you're ready, walk me through how you'd approach this.
Do you know anything about CI/CD?
Explain how you would apply CI/CD practices to an end-to-end machine learning system. Cover source control, automated testing, data and model validation, artifact versioning, deployment strategies, rollback, and production monitoring. Keep the answer practical and distinguish software CI/CD from ML-specific continuous training and delivery.