Upstart Research Scientist Interview Questions
& Process

So I interviewed at Upstart, and it was honestly the worst. After my final onsite interviews, HR called me and said they were happy to offer me the job, but they just needed to do a reference check with a previous manager and colleague, and also set up a short call with their VP of ML. That call with the VP was supposed to be 30 minutes, but it ended up being an hour for no reason. During the VP call, she asked some really weird questions like 'What is your dream?' and about my ideal career path. When I said I wanted to learn, do passionate work, get good mentorship, and grow with clear promotion paths, she got weirdly defensive and pressed me on why promotions were so important. The mood totally changed. She asked a couple of stats questions after that, which I think I answered okay. Then, a few days later, HR called again, sounding surprised, and said the team decided not to move forward with an offer after all. He said he was still figuring out why but apologized and admitted he was as shocked as I was. It's confusing because if the VP's vibe check was the deciding factor, why do a reference check first? It didn't make sense to do the reference check and then suddenly pull the offer.

The first round had two problems, one math/stats and one coding. If they thought I did well, I would have moved to the next round. I didn't pass the first round even though my answers were right. Maybe they care about how you explain your answers, or they want perfect AI answers?

The interview process overall was quick and pleasant. They asked a lot about probability, statistics, and machine learning, plus coding abilities. There were two screening rounds initially, both pretty much the same, testing probability and stats knowledge. The final part was a virtual onsite with two technical rounds and one other hour-long round. The technical rounds went a bit deeper into the same topics as the screening rounds.