Just joined FB as SWE in ML. 10 years of somewhat unrelated experience (not SWE) and PhD in CS before that. Base/RSU/BonusPct/StartingBonus/TC. TC doesn't include starting bonus. In chronological order of offer. Dropbox, IC4. 210/600/20/0/402 FB, E6. 225/1M/20/100/520 Google, L5. 190/920/15/100/449 Startup offers: base/equity(as options). The strike price for all options was close enough to zero —AI startup in Reinforcement Learning, Series A, no revenue. 125/1% —Ads startup, profitable, 200M revenue run rate, Series A. 225/0.08% —Cloud startup, 100M revenue run rate, Series D. 220/0.1% I didn't negotiate any of these. They were all the first offers presented. Google knew FB's numbers. I should've had a better strategy. Blind helped me a LOT in my job search. I learnt how much I need to leetcode (I thought 20 solved was a lot). Learnt about Engg levels, and that they matter (I came from a place with no levels). Learnt about expected comp. Learnt not to go work at a startup. Hope this post helps someone.
OP, how did you prepare for the interviews?
Congrats!
Nice
Congrats
Congrats!
Thanks guys! I'm not qualified to advise on prep. I didn't know what Leetcode was or what a System Design interview was 6 months ago. I've also never been an interviewer in a leetcode style interview. I just followed what folks here recommended (leetcode mostly), and got lucky.
What kind of work were you doing previously?
I worked on ML-like stuff but in a very different domain (not tech). Sorry, being a little vague here for anonymity
Congrats. Are you IC or manager? Is pay higher for ML swe vs general see?
IC. I don't think SWE comp changes based on whether you work in ML. But I'm just regurgitating what I've heard on Blind.
Is that like an ML engineer role? How relevant were interview questions to LeetCode problems?
Interview was like SWE, but system design was replaced by ML focused interviews at Google and FB (but not at Dropbox).
Thanks! Did you work at a startup before this?
Congrats! Is the point of leetcode to give you practice solving challenging DS&A problems, or to familiarize yourself with specific problems in the hopes of being asked them onsite?
Both :)