How to Answer: "Tell me about a time you identified and fixed a memory leak."

“Our Node.js API service experienced gradual RAM degradation, leading to pod OOM Kills every 48 hours under high load. I triggered heap snapshots at 20% and 80% memory utilization using Chrome DevTools/v8 inspector and compared heap allocation timelines. The delta revealed millions of uncollected RxJS event subscriptions attached to a global event emitter without teardown handlers. I refactored the handlers to auto-unsubscribe using `takeUntilDestroyed` hooks, patched the unclosed connections, and added automated load testing using k6 combined with memory growth thresholds in CI to prevent regressions.”

“In our React application, users reported browser lag after navigating across complex dashboard pages. I used Chrome Memory Profiler to record heap snapshots before and after page navigation. I found that websocket event listeners were remaining attached to the window object because `useEffect` cleanup functions were missing. Adding proper cleanup unsubscribers dropped baseline frontend memory consumption by 60%.”

“During local testing, I noticed my Python background worker kept consuming more RAM over time. I used the `tracemalloc` library to track top memory allocations and found an unclosed database cursor accumulating cached records in a loop. I refactored it into a context manager (`with` block), resolving the issue.”