I applied through a recruiter. The process took 1 week. I interviewed at LinkedIn
Interview
Recruiter prep chat before phone screen was very friendly and informative, the chat was useful for preparing to the behavioral phone screen.
Had two consecutive 1-hour “phone” screen video sessions with very senior engineers - coding and behavioral.
I was impressed by how engineers (and recruiters) are proud of the company culture, heard similar things from an employee friend.
I was supposed to come on site but decided to accept an offer at another company, for unrelated reasons, so did not move forward to the on-site stage.
I applied through a recruiter. The process took 2 months. I interviewed at LinkedIn (Mountain View, CA) in Sep 2018
Interview
The recruiter contacted me, there were two rounds of phone call, very interesting people. The second round was a lot around machine learning with a large number of fast questions. Overall satisfied about the phone call. The on-site, feedbacks and everything very very slow. It can take months . The on-site is usually Silicon Valley one day interview. I had few smart people, interesting questions. Overall the day positive. The review and results of the interview totally random and based on the feeling of 1/2 interviewers. So very bad point in that sense. The company seems having very little confidence with recruiter asking you continuously where you apply, how much they give you and which position. I’m neutral to whether apply or not here.
Interview questions [1]
Question 1
Usual ML staff (collaborative filtering, coding questions, ML design)
I applied through a recruiter. I interviewed at LinkedIn
Interview
The interview process was very well organized. Everyone showed up on time, and interviewers were extra friendly.
Second offer I get from LinkedIn. First one was several years back. So, the second time, they skipped the phone screening.
Interview questions [1]
Question 1
Design a recommendation system. The recommendation system can be used for multiple purposes (PYMK, JYMBII, etc). The system was discussed from both a ML and a systems point of view. That was the main question, and there were two coding questions, and there was a question on how would you respond to system failures.