## Scenario A team must choose Python, Java, or C++ for a new production service. Compare the languages and recommend one for each of these workloads: 1. a small internal automation service that integrates many APIs; 2. a long-lived, high-throughput business service with a large team; and 3. a latency-sensitive component with strict memory and runtime control. Explain when the default recommendation should change because of team expertise, ecosystem, deployment environment, observability, or correctness requirements. ### Constraints & Assumptions - There is no company-specific mandate. - Discuss total engineering cost, not benchmark speed alone. - Separate language guarantees from properties of a particular framework or runtime tuning. ### Clarifying Questions to Ask - What latency, throughput, startup-time, and memory limits matter? - Which libraries and operational platforms must be reused? - Who will own the service and what languages can they review and debug? - Are native dependencies, hard real-time behavior, or sandboxing constraints present? ```hint Tie every trade-off to the workload Runtime speed, type systems, garbage collection, binary deployment, library breadth, and developer iteration matter in different proportions for each scenario. ``` ### What a Strong Answer Covers - A reasoned default for all three workloads rather than one universal winner. - Python's iteration speed and ecosystem alongside runtime and typing trade-offs. - Java's managed runtime, tooling, concurrency ecosystem, and garbage-collection considerations. - C++ control and performance alongside memory-safety and operational complexity. - A decision process that includes profiling and a small representative prototype when uncertainty is material. ### Follow-up Questions 1. When would a mixed-language architecture be justified? 2. How do startup latency and steady-state throughput lead to different choices? 3. Which evidence would make you reverse your initial recommendation?