Whenever you're ready, walk me through how you'd approach this.
Design a bus management system.
Asked in the Round 2 stage.
Discuss how machine learning could support bus demand forecasting, vehicle allocation, route or schedule decisions, and operational monitoring. Clarify requirements before proposing an architecture. Cover data pipelines, online versus batch processing, model choices, evaluation, feature freshness, training-serving consistency, and failure handling. Explain how the system would operate when predictions are stale, unavailable, or incorrect.