Your question is Distributed KV Cache Architecture. 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).
Walk through the architecture of a distributed caching layer designed specifically for storing KV caches in multi-turn LLM conversations.