I have emailed a recruiter, sent her my resume and she got back to me to schedule a chat. The chat with the recruiter was about cultural fit. She checked that I cared about Linkedin mission and its core values. Nothing hard if you genuinely want to work there. After that I was sent a takehome challenge to be completed in 3/4 hours. As others were saying, if you prepare on the book "a collection of data science takehome challenges", you will be good. If you don't have time to go through all of them, just focus on the one related to predicting employee salaries... Before applying, I would suggest to practice on that book challenges and save R or Python scripts to then reuse them. That will save you a lot of time. Also, under time pressure, it is going to be so much easier if you have the main scripts ready. After the challenge, I went for the onsite. I was really scared, but it was actually easier than I thought. There were some probability puzzles, but really simple if you have a basic understanding of probability and if you prepared on the heard on wall street book. There was also one question where I had to describe how to implement and test a model to catch fake profiles. But the interviewer didn't ask me to go too much into details. It was mostly about which features I think would matter. I think they wanted to see if I had product sense. One day after the onsite, I got the offer.