It was a 3 rounds interview. First round was a chat about the role and your background, second was about presenting your work explaining how it relates to data sciences . In the third round you have to propose a solution on their data science problem they have within the company. I enjoyed the second round as I was speaking with people who gave me relevant remarks and questions on my work. The third round was with senior management people including a data scientist who presented him as experienced. As I did not know their products I asked a lot of questions (apparently too many) to be able to propose a solution. Then, without going too much into details I proposed a solution using recent NLP techniques. I wasn't really challenged on this solution because I believe the interviewer didn't have a clue about what I was talking about. They finally decided that my solution wasn't good enough without explaining why. Without clear and fair way to evaluate people it seems very much biased type of interview and then against the idea I have about data sciences.