Model Lifecycle
Domain: MLOps
🎯 Motivation
The full life of a model: conception -> data collection -> training -> evaluation -> deployment -> monitoring -> retraining -> retirement.
📋 Concrete Example
Lifecycle: v1 trained (Jan) -> deployed (Feb) -> monitored -> drift detected (May) -> retrained -> v2 deployed (Jun) -> … -> deprecated (2027). Full journey.
🔗 Analogy
A product lifecycle - design -> manufacture -> launch -> monitor -> improve -> relaunch -> discontinue. A model lifecycle is the product lifecycle for AI.