Your question is Monitor a Retrieval Ranking Stack. 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 are responsible for a large recommendation or search system with a retrieval stage and one or more ranking stages. The system serves live traffic, retrains regularly, and depends on both batch and real-time features.
What failure modes would you monitor in a large-scale retrieval and ranking system?