Tell me about a time you had to pivot mid-project.

In my last internship, I was building a dashboard component for a customer support tool, and the original plan assumed we could query historical data directly from one service. After I got the first version working, I noticed the data was incomplete for a lot of real cases, so I brought examples to my mentor and we confirmed the approach wouldn't be reliable. We shifted to a simpler version that used a smaller data set we knew was accurate and postponed one of the filters that depended on the missing history. I reused most of the UI work and updated the tests instead of starting over. We still shipped the dashboard for the main use case that sprint, and later the team used my notes when they expanded it once the data pipeline was fixed.

At a small edtech startup I was asked to add a real-time chat so students could ask tutors questions during lessons. About halfway through a two-week sprint, our compliance lead flagged that we couldn't permanently store messages for users under 18 and needed parental consent, which made my original storage design unusable. I suggested and built a pivot: the chat UI stayed the same, but messages for minors became ephemeral on the server, I added a simple consent modal, and I switched the backend to tag sessions instead of keeping full transcripts. I also wrote a short QA checklist and walked the product manager through the trade-offs so everyone understood the safety implications. We shipped the feature on time with the new restrictions in place, and the team appreciated having clear documentation for future policy updates.