Hello, happy Thanksgiving everyone! I have an upcoming interview with Google for the Research Intern position. I searched about the format of the interview but couldn't find anything relevant. The recruiter told me that the interview will be based on my research and there won't be any programming questions. However, this is too broad, can anyone give me some insights and tips? Thanks all! TC: Broke grad student
Do you know who is interviewing you? Would be good for you to find out that person's research background and interests. Researchers usually look for interns with similar skills and research interests as theirs. Just be yourself during the interview. Don't try to fake or make up stuff.
I can't see the interviewers, I think it will be general and team matching will be done after the interview feedback.
That's impossible. Research is not software engineering. Research teams usually require people with specific skills that fit their research domain. They can't just pull a random person out from a pool of available intern candidates. For example, you can't expect someone who does research in NLP to do research in network security. The skills just don't fit. This sounds strange. Check with your recruiter.
Hello! How was the interview for you?
Tech Industry
Yesterday
2882
The job market is absolutely brutal right now
Tech Industry
Yesterday
1531
I am starting to think Chinese interviewers currently fail non-Chinese candidates on purpose.
Tech Industry
Yesterday
2349
Crossed a line with my boss
Tech Industry
Yesterday
922
Update: Trans Coworker Stealing Breast Milk
Software Engineering Career
Yesterday
1350
Cleared Amazon onsite, but lowballed.
The research interview is pretty broad at Google. It could be something like asking you to prove a theorem in your paper or drilling you on the work you've done (justifying your assumptions, motivation, etc). They may ask some questions adjacent to your research too. Another variation is to ask an ML system design question (how would you design a bot detection algorithm for twitter, etc). Or present a hypothesis and ask you to design experiments to test it. The best thing to do is to know your research inside out, review relevant concepts of your domain (ML/DL etc). Maybe skim the interviewer's/team's work you know who is interviewing you.
Thanks, it is really helpful. I am not exposed to test design/data science related parts for instance hypothesis creation etc. However the rest makes sense. Thank you so much!
By the hypothesis I mean more like research questions and what experiments you'd run - shouldn't be much of an issue, it's simply brainstorming research ideas.