Your question is Root Causes in a Failing ML Pipeline. 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).
Problem scenario: given a detailed description of many things going wrong in an ML pipeline, answer about 10 questions on the root causes, how to fix the model, the tradeoffs, and how to explain the situation to stakeholders and higher-ups.
Asked in the 60-min problem scenario exercise stage. Tagged as a hard question; about 10 sub-questions. Reported follow-ups: What are the root causes?; How would you fix the model?