Hi Guys, I have ML software engineering interview (virtual onsite) with Google in two weeks. It will involve two rounds of ML design, two rounds of coding and a behavioral round. For the ML design round part, I have been prepping for ML system design questions such as designing feed ranking algo or estimating CTR etc. I am curious if ML design round will involve coding and derivations as well. Do I need to prepare for implementing ML algorithms from scratch and be able to derive ML algorithm from scratch? The recruiter was not sure if there will be some form of coding and derivation in design round. Is it good idea for me to start prepare for coding ML from scratch ? I appreciate any inputs and any tips for success? Thanks. tc :peanuts
Not sure about ML design session. My guess is that it is as you expected (like desiging home feed, how you build/deploy/scale it). But I had a coding ML session. After going through one of my projects, he asked to code K-means. It is good to know high-level of how basic and common algorithms work. Then, it can be easy to show you can code it.
Code k means without using the direct function, as in code k means from scratch? 🤔
Yeah, I was asked the same question at TripAdvisor phone interview, you have to use numpy functions like argmin
🍿
Hi, Can I dm? I’d like to know the process. I’ll be interviewing with them soon.
Op, please share the interview experience. Thanks!
Hi, could you please share your experience? I am interviewing for the same role in the next 2 weeks. Would really appreciate your help. Thanks!
Hi can I dm you?
Hi, can you please share your interview experience?
Tech Industry
Yesterday
1143
Update: Trans Coworker Stealing Breast Milk
Tech Industry
Yesterday
3489
The job market is absolutely brutal right now
Tech Industry
Yesterday
1729
Did this candidate cheat?
Tech Industry
Yesterday
2938
Crossed a line with my boss
Tech Industry
Yesterday
2052
I am starting to think Chinese interviewers currently fail non-Chinese candidates on purpose.
For the ML design, you will typically NOT be asked to code but it can happen. You can’t fake it though. You need to have worked on ML projects to be good at answering an open-ended question in the lines of « We have this data, we have this problem, how can ML solve it ».