Your question is Model Selection Tradeoffs. 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're comparing several candidate models for a supervised learning task and need to pick one for production. The team wants a clear rationale for why one model is better than the others, not just a higher validation score.
How would you explain the trade-offs behind your model selection?