Your question is Handle Feature Drift in Production. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you handle feature drift in a production environment?
Explain how you would detect changes in feature distributions, distinguish harmless seasonal variation from harmful drift, and determine when to retrain or roll back a model. Include monitoring, alerting, validation, and safeguards against training-serving skew. Provide production-quality Python that compares a reference dataset with a serving batch and evaluates whether drift should trigger action.