Several stages: - Zoom call with AI engineers where went through a public repo of mine, which they asked me provide a week before. -On-site -- Research presentation: 1 hour presentation about my own research, they were very inquisitive about my research both during the QA session and in later interviews. -- Whiteboard interview with 2 engineers (algorithms + data structures): Started with basic questions, then moved towards more open-ended questions (e.g. How would you make this faster?). -- Whiteboard interview with 2 engineers (ML knowledge), some math questions and going through pytorch code on a screen. -- 2 1:1s with a Senior AI scientist (ML knowledge): Questions about both my research and general ML knowledge. -- 1:1 with 1 AI scientist (research and coding, behavioral), chatted about our research, and past experience. -- 1:1 with VP of engineering (past experience) Overall, I found everybody I talked with to be nice people with a positive attitude and passion towards what they are trying to accomplish as a company.

There was one HR call for 30 mins and a 45 mins background screening with the Tech lead. Finally, there is a technical interview on machine machine. But it trurns out to be algorithm question under the shell of data loader.