ML training pipeline

flowchart LR
    raw[Raw data] --> clean[Clean + label]
    clean --> split[Train / eval split]
    split --> train[Train model]
    train --> eval{Metrics pass?}
    eval -->|no| tune[Tune hyperparameters]
    tune --> train
    eval -->|yes| registry[Model registry]
    registry --> deploy[Deploy endpoint]
    deploy --> monitor[Monitor drift]
    monitor -->|drift detected| clean