Interview process included a live technical screen with the CTO. The task was a simple Califarnia postcode format validator. Google, Stack Overflow, and AI tools were explicitly allowed. The stated goal was to produce a working solution within around 25 minutes and be able to explain/defend it. The interview was less about the difficulty of the coding task itself and more about how the candidate uses AI under pressure. The interviewer frequently challenged decisions during the exercise, including how the AI prompt was written, whether the source material was being used correctly, and whether the generated solution was too complex for the actual requirement. My read is that they are looking for engineers who can strongly own the problem definition, push back when needed, and use AI only as an accelerator rather than relying on generated output. The feedback was direct and quite intense. It may be a good fit for candidates who enjoy high-pressure, outcome-driven environments, but it could feel adversarial if you expect a more collaborative live-coding format.