Your question is Approach to Underperforming Models. 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).
You've shipped a model that looked acceptable during development, but it is now underperforming when used in practice. Your team wants a structured way to diagnose the issue and decide whether the problem is data, metrics, thresholding, or the model itself.
How would you approach a problem where your model is underperforming?