I applied through a recruiter. The process took 2 months. I interviewed at Google (Mountain View, CA) in Jan 2018
Interview
Recruiter found on LinkedIn followed by one phone Screen on DS and Algos. Onsite had 5 technical rounds (4 coding, 1 ML) + lunch. There is heavy emphasis on writing whiteboard code which is complete in all aspects. My word of advice is don't sit down until you write down the complete code on the board. Rapport with interviewer and other things don't matter as much as much as awesome whiteboard code. One ML interview was very light. Interviewers were nice overall.
I applied through a recruiter. I interviewed at Google (New York, NY)
Interview
4 rounds (online) - 2 coding, 1 ML systems, 1 behavioral. Coding were standard leetcode style. ML system was expected. Behavioral was also the standard questions. All went pretty smoothly
Screening interview is about a coding problem. It was a 45 mins interview. I had no question about machine Learning. It was only about coding and software engineering. The mistake I did is that I spent too much time explaining my approach on the first question. While the interviewer had more questions for me.
Classical google interview process;
HR phone call,
Algorithm interview and other technical interviews.
Good to have a feedback after every step from hr. However unnecessarily difficult algorithm problems are annoying, after 10 years they are still doing same, no improvement.
