Your question is Real-Time AI Analytics 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).
Design an architecture for a real-time analytics platform using AI at Bloomberg. What components would you include?
Explain the end-to-end data, ML, and serving architecture, including ingestion, feature computation, model training, online inference, retrieval, ranking, and monitoring. Clarify the missing workload, latency, freshness, security, and availability requirements before making capacity assumptions. Address how the design handles delayed labels, model and feature drift, training-serving skew, Bloomberg data entitlements, and degraded dependencies.