Skip to content

Latest commit

 

History

History
70 lines (37 loc) · 4.5 KB

faqs.md

File metadata and controls

70 lines (37 loc) · 4.5 KB
Raw

List of common questions

Question for people have no working experience in DS/ML

  • What is the strategy/study plan to get a job in DS/ML? I don't know. There are many ways to make up for knowledge gap. For a competitive field like ML/DS, it's common for people to do the following: 1) get a Master in DS/ML 1) Try career-focused program like insight fellows 2) interview query.

  • Common questions about my interview experience.

Questions about focus areas

  • What is the MLE interview process? This document described the process very well.

  • Did you get questions around PyTorch, TensorFlow and other frameworks? No.

  • Which language should I learn for MLE? Python is useful for both interview coding questions and actual work (tensorflow, pytorch, scikitlearn, pandas etc).

  • Do I need to implement ML algorithm from scratch? Yes, see details in study guide

  • Do I need to solve LC hard? Yes. I got about 5%-10% questions are LC hard.

  • Were you asked about cloud technologies, containers, orchestration, etc? No, but it's a MUST have when you present system design interview.

  • For ML theory, was it more focused on classical ML or DL? Both.

  • Were you expected to be up to speed with recent cutting-edge research in ML? No, nice to have. At the minimum please read about Embedding.

  • Is it ok to be a generalist or are you expected to have a speacialisation in Vision, NLP, etc? It depends on roles and company. In general, it's not required.

  • Were there any questions around various Data Engineering tools? In general, No. But a lot of companies expect you to know how to handle big data with Spark etc.

  • Did you get probability and statistics questions? If yes, what kind? Yes, Bayesian. And nobody care if you know Gamma and Poisson distribution.

  • Were there any ML System Design or general system design questions? Both.

  • Do the interviewers ask indepth questions in ML like explain Adagrad or XGBoost in detail? Yes, but do not try to memorize the details, see how to study

  • Do I need to read any research papers for the interviews in top companies? No, nice to have.

Questions about before/during on-site interview

  • What to do one day before the onsite? Relax, do not study. Go to sleep early. If you have to fly, please do not fly on the onsite day.

  • Do I need to customize my prepraration for company? I do that and found it's extremely helpful. I spent one hour or two on glassdoor to get the overview of the interview. I also review interviewer's Linkedin thoroughly.

  • Do interviews need to like me? Yes, it helps.

  • What to do when you black-out and have no idea how to answer? Take one step back, breath deeply, do not think about anything, just give 30 seconds or so to calm down. If you know the answer, it will come to you.

  • What is the right way to answer technical questions? Think about interview's intentions then answer to the point. Try to stay precise, concise and to the point.

Questions about logistics

  • How long does it take to prepare for interviewing? It depends, 3-4 months is a stretch but doable.

  • Do you recommend doing a few interviews are startups as practice before moving onto the big tech companies? Yes, after 2-3 startup companies there is diminish of returns.

  • Do you have any negotiation tips you can share? Two simple rules: 1) Do not specify the TC number first. 2) Have competing offers for negotiation.

  • Did you get many questions about your past experience? if yes, how did you navigate not being able to share sensitive information? Yes, tell interviews, they understand that.

  • Do you think non-CS background affects my chances in any way? Not to my knowledge.

  • How should I go about preparing? See plan

  • Can you help in how to get through the in-depth ML questions asked in the interviews? What sources did you follow for that? See plan

  • Can you provide me mentorship on what to prepare and how to maximise my chances? Yes. Send me an email to helppreparemle@gmail.com.

Notes

  • If you're interested to learn more about paid ML system design course with more examples, click here.
  • If you find this helpful, you can Sponsor this project. It's cool if you don't.