The interviews included demoing previous work I'd done (public product) and answering detailed questions about the technologies in my area of interest. Algorithmic coding on the whiteboard was also part of the process. Ultimately, I think only the algorithmic coding was significantly considered, based on the internal recruiter's feedback about all candidates (though they wouldn't provide information specific to me).

The problems presented were atypical to coding we do in computing in general, much more aligned with what I studied as an undergraduate. Big-O analysis and scaling were prominent. I believe there is a disconnect between what they tested and what engineers in the positions I was focused on do on a regular basis, with neither necessarily being a good measure for the other.

I found all the interviewers very friendly. It was as if they wanted me to succeed and would ask helpful questions when I was off base. A friend (also interviewing there) and I wondered if their model is that everyone gets the answer, and the measure is just how much help they provided.

They asked that we keep the interview questions confidential, and I will respect that.