They ask various questions about your general research and the problems you are currently working on. They also ask various hard technical questions related to your research area, so they can understand if you have a strong grasp of the research literature. They are also interested in coding projects you have worked in the past (especially if you apply for a very technical project). You basically have to showcase that you are a great hacker + have great research skills (through papers, talks, etc).
Interview questions [1]
Question 1
Describe your contribution in one of your main papers, and how it compares with the state of the art. Then a lot of discussion on it, why you did things a certain way, etc.
I applied online. The process took 4 weeks. I interviewed at Microsoft in Mar 2014
Interview
A manager contacted me from MSFT and informed me that they are looking for research interns. I had two interviews. One coding interview and one data science interview. The coding interview was not very challenging however the data science interview was hard!
I got an e-mail about a week after my last interview with an offer.
Interview questions [1]
Question 1
I cannot disclose any explicit information about the questions. However I can suggest people who are interviewing for data scientist-like positions to do a quick review of popular statistical techniques and be prepare to answer questions like "So why this method work?" or "Can you sketch a proof of why this method work?". Be creative and do not be afraid to answer questions with out-of-box thinking. There are many good sources online for interview questions for data scientists. You can review the most common questions in a day. So do it! It'll pay off.
Phone interview then in person interview. The phone interview was brief and just went over resume. The in person interview was more in depth and asked standard behavioral questions. It also included questions about what to do in certain situations. Interviewers were nice and helpful.They asked many questions about personality, strengths and weaknesses.