Micro1 Software Engineer interview: AI assessment instability

My interview opened with an AI-driven conversational segment before moving into technical tasks. It felt more like an assessment than a human conversation. The sections tested whether I could reason clearly, explain tradeoffs, write code, and debug under time pressure.

I worked through live coding and debugging, code-optimization tasks, and API questions. The questions were not vague in the way some AI interviews can be. They centered on execution and clarity: how I approached a problem, reasoned about bugs, and improved code efficiency.

The technical work was the most concrete part, but the overall run still felt unstable. In another attempt, the AI raised flags without a real obstruction and never produced a result. In the more complete session, I could make progress on the tasks, though the flow remained impersonal. I did not have a positive outcome. What stayed with me was how much the experience depended on the platform working properly while I tried to show my skills.

Location: Bengaluru. Overall feedback: negative. Offer status: no offer.