Your question is Designing for High Availability. 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 a highly available, distributed system to handle high request volume for Seven Ai's agentic security platform?
Explain how you would clarify traffic and availability requirements, partition the online and offline paths, and serve security decisions within a strict latency budget. Cover the request path, model inference, data storage, scaling, observability, degradation behavior, and evaluation strategy. State assumptions explicitly where production traffic, catalog size, or model details are unavailable.