After the OA I went straight to the virtual onsite. It was the new vertical consumer team. There were two rounds in total: the first was AI craft coding and the second was knowledge. Each round was one hour, with a 15-minute break in between.
The AI craft coding round did not ask the workflow engine question that people have reported on the forum. Instead it was a very simple dataset-building task. HackerRank gives you 5 small datasets related to orders, ids and stores, and you need to turn them into a training dataset: pick features from what is already there and add labels (if I remember correctly, whether an order was placed or not). The datasets can have various small problems, such as duplicates and missing values, and you have to deal with them. You use Codex inside the IDE at medium effort, write prompts for it, talk with the interviewer a bit, and interpret the results once you get them.
The second round was supposed to be a trial run of testing domain knowledge, but it felt like a completely behavioral round. They ask all kinds of BQ questions and you answer by tying them to your own experience. The experience was mediocre. If your answer is research-related, they may ask some technical details.
After the interview I went straight to nudge the recruiter, because I have other things going on, and they promised to try to get me a result within two days.
Discussion
Loading comments…