Hey Blind folks, Thank you for helping me get here. I prepped intensely for about 3 months and switched to Google from a mid-size unicorn. This is for ML engineer IC with ~10 years of experience in the Bay Area. I thought I should share my journey to give back. I interviewed around 10 companies and got at least 6 verbal offers. Dropped some tier 2 companies due to them being slow. New TC: ~600k/year in 2022 Spread over 4 years: ~480k/year (due to front loading and sign on) Previous TC: ~270k/year (including option gain with buy backs) Offers from Google, Meta, and Uber really helped bump up the numbers. I provided redacted written proofs to Google to get their offer bumped up. I studied leetcode the least (around 2-3 weeks) relative to others. Probably over 10 years I did around 150 leetcode problems (make sure to analyze the patterns). Also, very important to display your thought process. Key is to verbalize methodical thought process and solution is optimized for use-case. I sometimes didn’t get written solutions but by walking through thorough thought process, they gave me more coding rounds or give you the benefit of the doubt and move you forward. Mostly spent time studying ML concepts, ML system design, and general systems design. Some material that helped came from Educative.io / YouTube videos / Alex Xu’s book, etc. Also real world experience building, designing, and discussing systems helped during interviews. Also check out engineering blogs from companies (confluent / Uber / Airbnb / google / fb / Netflix / etc). If you have time you should probably read designing data intensive app book (I didn’t have time). During design rounds it’s important to mention the various concepts you might apply or consider even if you can’t use it due to interview time constraints. Because they can’t assume you are aware of it unless you mention it. They also ask follow up questions on them if they think it’s important. On-site interview rounds looked like 2 coding / 1 ml system / 1 general systems design / 1 behavioral for most big companies. Sometimes some companies do 2 ml systems on top of general systems design or no general systems design. When studying for ML, it’s very important to understand approaches in building ranking models in production as it can be used in the context of feed / recommendation / search, which are what most internet companies get ROI on ML. Also review your specialization specifics as well. - Google moved very fast for me. After on-site, I got written offer in about 3 weeks. - Google do low-ball at start, but they were willing to pay and get pretty close to Meta (but 5-10% less). Uber helped get Meta go higher. Uber seems to pay more than Meta to compete. Hope this helps! #tech #machinelearningengineer
I have my on-site with meta. Can I dm you please?
Congrats. Was it L5 at google? Did google have a general system design round as well? If no, did they do 2 ml designs?
1 ml system and 1 general system.
Can I dm you? What was your final offer from meta? Also, breakdown for offer from google?
Meta came to be around 580k first year due to ~90k sign on. 200k+ base 800k+ RSU 33-33-22-11 and ~70k sign on for google.
Feel free to DM.
Good advice about ranking models. I have lots of exp in ML but not in ads/ranking. Any good pointers specifically for them?
https://youtu.be/0nu83yWqnNQ seems good. Educative.io has ML engineer interview course. That also helps get some basics down. Learn about featurization for pointwise ranking, e.g. embedding. Know the basics of pairwise and listwise ranking. I think Airbnb has some good blog posts about ranking as well.
Thanks!
Do you need to show counter offer in written proof to Meta? Most of the times all are verbal offers
I didn’t need to.
Congrats. Welcome.