Your question is Scaling Backend Services. 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).
What are the primary considerations when scaling out a backend service?
Discuss how you would reason about workload characterization, horizontal scaling, state management, data access, reliability, observability, deployment, and ML inference concerns. Address how you would validate capacity assumptions and handle degradation when dependencies or model-serving infrastructure are unavailable.
