A company deploys a production feature pipeline update, and the next inference batch shows a sharp drop in non-null values for a key risk score. The batch transform job completed successfully. What monitoring workflow would best prevent bad predictions from being consumed?
Your Answer
Explanation
FREE AI FEEDBACK
Guest checks show correctness and the answer key. Log in to save history and unlock evaluator notes.
i
Review the explanation and try similar questions to strengthen your understanding.
Issue reporting
Report an issue
Clara
How can I help?
Search AWS AI Certs Reviewer
Search across lessons, syllabus topics, provider capabilities, and certification questions.