Your question is CAP Theorem and Distributed Systems. 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).
Explain CAP theorem tradeoffs and how you design distributed systems under those constraints.
Discuss how partition tolerance affects consistency and availability, and explain why a practical system must choose behavior during network partitions. Cover data stores, replication, quorum or consensus protocols, failure handling, observability, and how the tradeoff changes for ML pipelines, feature stores, and model-serving paths. State assumptions clearly and distinguish control-plane requirements from latency-sensitive data-plane requirements.