ML Observability

Tracing one request through features, model version, prediction, decision and outcome; logging what is necessary and safe; and debugging an incident from the business metric down.

ML Observability
▶ lab

A healthy model server can serve a wrong model indefinitely. Observability for a model means tracing Request → Features → Model Version → Prediction → Decision → Outcome, and watching signals service health does not have.

Q · The serving dashboards are green and the model is wrong. What signals does a model need beyond service health, and how does one prediction get traced to its outcome?
Prediction Logging
▶ lab

The prediction log is the monitor's input, the incident's evidence, the outcome join's left side and the next training set. Log what is necessary, reference what is sensitive, and decide retention before the first row.

Q · What does each prediction record need to contain for monitoring, debugging and retraining, and what must it not contain?
Tracing a Prediction
▶ lab

One request id, from the click in the frontend through the backend, the feature service, the model server and the decision, to the outcome event weeks later. Each hop records something specific, and a hop that drops the id is where the next incident becomes a guess.

Q · What does each hop between a user action and an outcome record, and what becomes undiagnosable at the hop where the correlation id is lost?
ML Incident Debugging
▶ lab

Conversion fell after a deploy. Investigate from the business metric down — prediction distribution, model version, feature values, feature pipeline, raw data — in that order, and do not retrain until the cause has a name.

Q · The business metric dropped after a model deployment. In what order do you look, what does each layer rule out, and why is retraining before diagnosis the wrong move?
Model Postmortems
▶ lab

A model incident postmortem records the assumption that broke, the signal that should have fired, the label delay that hid it and the test now added — and never concludes that "the model" was at fault.

Q · What does a postmortem for a model incident have to record that a service postmortem does not, and why is "the model was wrong" never a root cause?