Domain 5 of 6

Monitoring and Incident Response

The stage teams skip. Drift, monitoring that watches output quality, what a rollback criterion is for and how to handle the day a system produces something that reaches the press.

Concepts

~12%

Of the exam

Practice questions

Concepts in this domain

01AI incident responseWhat a team does on the day its system produces something it should not have, the decisions worth making months before that day, and when the law requires the incident to be reported.02The machine learning lifecycleThe stages between a business question and a model in production, why the work does not stop at deployment, and where each cloud vendor's tooling sits along the line.03Transparency and explainabilityWhy an accurate model nobody can explain is sometimes the wrong choice, the difference between the two words, and what a model card is for.04Cost and consumption controlsWhy unbounded consumption is a safety problem before it is a finance one, where the money in an agent run actually goes, and the limits that decide what the worst case looks like before somebody else decides it.

Try a question from this domain

A live assistant has told several customers something incorrect about their entitlements. The system is up and almost every other answer is fine. What should happen first?

AStop the harm, by switching the feature off, restricting it or routing everything to human review.
BReproduce the failure so the cause is understood before anything changes.
CDraft a customer communication for approval.
DChange the prompt to prevent the specific wording that caused it.

11 questions on this domain.

One per page, with a worked explanation.

Start the set