Finished a virtual onsite with Meta for an E4/E5 SWE AI role. Here is my experience: Coding 1: Finished 1 easy/medium level bit manipulation question. Had a good discussion with interviewer regarding cache optimizations. Interviewer seemed happy with solution. The next problem was an open ended linalg coding question. Was able to code up initial implementation + 1 step optimization, however it seemed my interviewer wanted me to solve it with a diff design (ran out of time). I would say I passed this round but could have done better with the second question. AI Design 1: Open ended design question with many areas to dive deep. Was able do discuss many of the high level components of design. Interviewer seemed to have similar experience in the stack as me: was able to answer most questions/followups with confidence. OK interview overall, not exceptional, not bad. Interviewer was nice, seemed to look forward to working w me. AI Design 2: This might have been my best interview, was able to answer all of the questions/follow ups my interviewer had prepared. Interviewer mentioned I had finished all of the follow-up Q with ease. Good vibes overall. Follow ups were pretty high level, did not have to answer any specific/hard questions. Interviewer seemed happy with my answers. Coding 2: was given a linalg easy level problem that is on LC. Had not seen this problem before. Discussed different ways to solve. Interviewer nudged me into the right direction and was able to find the solution and code it up quickly. Wrote optimal CPP solution. I have my doubts if the interviewer thought I needed a hint to find the solution. Q2 was a medium level question, I had seen something similar. I quickly pointed out a trivial way to solve using sort. Interviewer wanted a better solution. I knew how to solve this problem using optimized sorting algo so proceeded to implement that. Tested my solution for both problems with examples. Interviewer pointed out few errors while implementing. Interviewer was poker face, I even think he did not like me. Behavioral: STAR method. Prepared different situations and tried to adapt my experience to answer the behavioral Qs. OK interview. Edit 1: Received Offer for E4
I see Meta, I vote no offer.
^ this OP, take the results with a grain of salt, half of people on blind dream of working at Meta and that tends to distort results…
Eh I don't vote no because I dream of working at Meta, quite the opposite actually. I vote no because Meta interviews are just too weird, they want perfect LC robots. Any doubt in mind in a Meta interview usually results in a reject. That has been my experience multiple times, ymmv. I'll be very happy if OP makes it!
yeah any hints they think required probably no offer. I solved multiple LC m level questions with Spacex across two groups, and went to take home project stage, and was asked follow up questions on project. No offer lol I think HMs see layoff announcements from major companies every day and like to play the field...
Keep us posted, GL! May the luck be with you
Sometimes you underestimate yourself, rating yourself an ok interviewee but things may turn out different. Good luck!
Yoe?
Good luck!
Congrats on the offer, OP! Looks like two of your coding questions involved linear algebra. Is this because you were interviewing for an AI role? Were these tagged “FaceBook” on LeetCode?
Work Visa
2h
341
How many of you in EB-2 India category think of going back to school for PhD and getting a bunch of publications?
Tech Industry
Yesterday
2913
1 vs 5 Million - no lifestyle change
Ask Blinders
Yesterday
727
Why is our country owned by Israel? I don't want my tax dollars fund genocide. How can we stop this nonsense?
India
2h
378
What do vegetarian Indians eat for protein?
2-4 years ago this would have lower band e4 offer given the current market and the caliber of talent looking for opportunities I’m leaning to no offer