Your question is Data Drift vs Code Bugs. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
If a model's prediction accuracy declines after deployment, how do you differentiate between data drift and code-level bugs?
Explain a practical investigation using production prediction logs, reference training data, feature distributions, model outputs, and observed labels. Your response should distinguish statistical evidence of drift from failures in preprocessing, feature ordering, model loading, thresholding, or monitoring code, and should describe the validation checks and remediation steps you would use.