Describe a situation where you had to show bias for action in a tight deadline scenario

In my last internship, we found a bug two days before a demo that caused one part of the app to fail for new users. My first step was to reproduce it reliably and check with my mentor whether this was the highest-priority issue, because I didn't want to spend time on the wrong thing. Once we agreed it was, I focused only on the signup path, put aside a small cleanup task I had been doing, and shared updates as I narrowed it down. I found that a recent validation change was blocking one required field, fixed that path, and stayed involved through testing to make sure we didn't break existing users. We made the demo, and afterward I added a test for that case and documented the issue so we wouldn't hit the same problem again.

At my first full-time role after bootcamp, we discovered two hours before a scheduled client handoff that their CSV exports were silently missing rows. Instead of waiting for the senior on-call, I grabbed the logs, reproduced the problem locally, and found a timezone parsing change from a recent library update was dropping midnight records. I rolled back that one dependency change in production to restore exports, ran a few manual checks with real client data, and notified the client and my manager about the temporary fix and the plan for a proper migration. After the release I wrote a small automated test to catch the timezone case, documented the incident, and scheduled a safe upgrade path so we wouldn’t have to rush a similar rollback again.