The phone screen was coding plus a project deep dive. The coding was that Dasher Pay question. The interviewer was really nice: right after I finished writing the code, they just told me not to worry, they were passing me. For the project deep dive, the interviewer listened really carefully and asked a lot of questions.
The VO was four rounds.
Round 1 was AI coding. I think they have two coding questions for different levels. For senior or staff it should be writing two services that can interact with each other to complete a refund service (you can check earlier interview posts for the specific requirements). The key point of this question should be that the services you write can simulate a real refund scenario, but every interviewer focuses on something different. I think this round is testing whether you've already thought through the architecture of the whole system; the AI is just there to assist.
Round 2 was system design, that menu review question. This is probably where I failed. This question has a lot of functional requirements, and I wanted to get through them as fast as possible so I could get to the deep dive later. But the interviewer in this one kept asking about lots of data schema and API details, so at around the 45-minute mark I still hadn't started the high-level design. And I feel like a lot of those questions actually already counted as deep dive and should have been asked after I showed the high-level design, but there was nothing I could do, the interviewer just wouldn't let me get to the high-level design and the deep dive. So I feel like I spent a lot of the time showing mid-level engineer signals.
Round 3 was incident. This round was actually pretty interesting. I think if you've handled a fairly big sev before, you shouldn't have any problem.
Round 4 was BQ. This round was only 45 minutes, and their format is pretty unusual: it's rapid-fire Q&A. I think the HM who interviewed me was zoning out the whole time, because the questions were really broad but you were expected to answer each one in about 5 minutes, so you give a little context and then have to get straight to the point. The HM didn't ask a single follow-up either. One, I don't think they were listening; two, I think they were also short on time. Anyway, at the end they can just let the AI meeting notes sort it all out. But even though they were zoning out, this round should still have been positive.
Two weeks later I was told I was rejected. My feeling is that with them, if you don't get an update within 48 hours, it's basically over.
Discussion
Loading comments…