the 1st ML-case was about creation of recommenation blocks on home page - what data to use to train model, why it is important, bla-bla. It was given 3 days before interview. the 2nd case was given during the interview, it was about building ML model to predict in how many days from today (if any) the hotel will be fully booked for given range of search dates. What I did not like about interview process: 1) People's accent when they speak English. It was my 1st experience of interview in such intl. company, and I was a bit shocked about their accent. I really could understood 10% of what one indian guy tried to tell me. The same problem was with another french woman (firstly, I thought she was asking me smth in french) 2) I feel I have failed the 2nd interview with online ML-case because it is hard to figure out the solution without thinking at least couple of minutes. It was not a standard ML setup, you need to figure out how the data should look like, the target etc. However, those guys did not give me that time, instead they started asking me some irrelevant questions about how gradient boosting works, why it is better than RF etc. just wasting my time. 3) I do not know why, but all of them asked me the same question about my last project I worked on current job. So I told this story 3 times. 4) Finally, they did not provide me with any feedback after 1+week from the last interview. For me, it is the worst signal about company. I spent several days making presentation, meeting those random people, and they could not send me final E-mail.