I received a first response one week after submitting my application. After an initial lengthy period, things started moving fast. The screening call with the recruiter was trivial.

The second interview involved a set of Python and SQL questions with increasing complexity. The main point here is timing: you don't have to aim for the ideal solution, just something that gives correct output. You should also explain what you are doing. The interviewer was neutrally positive; there was no personal touch, however, I felt no one wanted me to fail.

The final interview was a full day, in-house. The questions were a mix of programming, SQL, and data modeling questions. You also had an opportunity to ask questions.

I believe the programming and SQL questions were of average complexity, and there are ample resources on the internet to help you prepare.

Regarding data modeling, this is a bit of a grey area. I believe I was asked questions that could be answered in multiple ways, and I had the impression only a certain aspect of the answer was accepted. I couldn't understand which part of the solution they wanted me to elaborate on in more depth. This could potentially be a weak point of the process and might lead to some random results in hires.

There is a focus on huge amounts of data. Think about how to pre-aggregate very large datasets into smaller tables that could, however, be re-used to answer multiple business questions.

I believe most of the interviewers were nice and pleasant. They need to put down some notes and remarks during the interview, which might be distracting.

In general, it was a good process with minimal frustration relative to the company's size. My only frustration was with data modeling. Should I ever decide to re-apply, I might still fail the last stage as I still don't understand what answers are expected from me, even though I have vast experience in this domain and have architected many big data solutions.

Design a dashboard to highlight a certain aspect of user behavior.