Your question is High-Traffic Caching Strategy. 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).
Discuss how you would implement a caching strategy for a high-traffic application.
Focus on how caching fits into an ML serving stack, including what to cache, where to cache it, how to keep it fresh, and how to avoid stale or inconsistent predictions. Keep your answer scoped to architecture, tradeoffs, and operational concerns.