Your question is AI Inference With Queues. 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).
How would you design an AI inference system that takes user inputs from queues and returns async responses, including rate limiting, database design, and API design?