Taro Logo
3

I keep failing my ML/Data Science interviews and I dont know why

Profile picture
Junior Machine Learning Engineer at Taro Communitya month ago

In the last month I had 5 companies I interviewed for. I made it to on-site for 2 companies and got rejected after first round for 3 interviews and i'm feeling so lost on how to get better or what I'm doing wrong

  1. a series A YC startup: they ghosted me after a first round which was a HM convo about my past experience. Didn't even send an email even after follow up

  2. Wayfair associate level role (asking for 1-2 YoE): passed OA. idk what happened i thought it went well but I got rejected after first round. It was a "case interview" for data science. Dont want to leak the exam on public forum but it was something along the lines of they said u have X data, what would you use it for? How to train a model on it? and a lot of follow up questions.

    I took a mock interview for a system design from interviewing.io and got passed at the mid level so im not sure why i got rejected here.

  3. a series B startup: passed OA/takehome. failed on site - 4x rounds (SQL, pandas, coding/pair programming, ML theory) I thought I did fine on everything except the SQL which honestly is not my storng suit. I did ok ok on it. I kinda fumbled on 1 question out of 6 questions of the ML theory round where they asked me a stats question (find sample size needed for calculating significance of an A/B test). But I think I did well on the pandas round and the rest of ML and coding/pair programming.

  4. a really really fancy AI startup hiring ML Scientist: I did a 4 hr take home which I passed and then a 5 hour onsite no DSA but really delving into ML research skills and system design and coding. I was totally unqualified for this (they wanted strong research/math skills) so im not surprised here

  5. Series B startup: Passed OA and I got rejected after the first interview the moment the HM realized I had 6 mos of experience he ended the interview right there

Didnt also make it past the phone screen for 2 companies. I presume they were looking for someone more senior based on the JD

70
7

Discussion

(7 comments)
  • 0
    Profile picture
    Tech Lead @ Robinhood, Meta, Course Hero
    a month ago

    #1 - They probably wanted more experience, and sometimes the recruiter just fumbles and does the YOE math wrong, giving a phone screen to a candidate that never had a chance.

    #2 - This is hard to tell. It could be a YOE thing (you say it required 1-2 YOE and you have 6 months). What I will say is that this is the problem with mock interview services: It's impossible for them to really match the real thing. And in this market, the real thing is insanely hard.

    #3 - Probably the 1 question you semi-fumbled? The bar is just so high in the current market.

    #4 and #5 - Unqualified as you mentioned.

    All of this definitely sucks, and this is simply the reality for junior engineers in this market. I wish I had some magic wand to wave to make your interview rejections go away, but I don't 🥺. In this market, it's just about having grit, setting realistic (i.e. low) expectations, and not taking anything too personally.

    It's not likely they'll respond, but you can also try asking for interview feedback.

    • 0
      Profile picture
      Junior Machine Learning Engineer [OP]
      Taro Community
      a month ago

      So I got interview feedback for #2 and it was bc I didn't answer the case study well enough and fast enough. I took too much time to think through/ask clarification questions that I didn't get thru the whole study in the 30 mins allotted. It looks like one of those cases where it's a role on a very specific team (notification personalization) and the questions were very tailored to that so you either know it or you don't

    • 0
      Profile picture
      Junior Machine Learning Engineer [OP]
      Taro Community
      a month ago

      I think the thing I'm struggling with is that the interviews for semi-late stage startups seem so personalized that its like really really hard to prepare. Like I need to be flawless in SQL, python, pandas, statistics, ML, DSA which is so hard so I'm feeling a bit lost on how to prepare well. Obviously just DSA itself can take 100s of hours to be flawless. On top of that having to be flawless in SQL and pandas and stats and everything feels impossible

    • 0
      Profile picture
      Tech Lead @ Robinhood, Meta, Course Hero
      a month ago

      That's actually reasonable feedback. It's rough advice to hear, but "Be faster" is very valid for interviews and always has been. The specialized nature of the interview makes sense too.

    • 0
      Profile picture
      Tech Lead @ Robinhood, Meta, Course Hero
      a month ago

      I think the thing I'm struggling with is that the interviews for semi-late stage startups seem so personalized that its like really really hard to prepare.

      It's just startups in general unfortunately. It won't feel as bad for early stage startups since their hiring bar tends to be lower (early stage means they have less hiring power and are more willing to take bets on talent).

    • 0
      Profile picture
      Junior Machine Learning Engineer [OP]
      Taro Community
      a month ago

      Thanks for the advice Alex, it helps a lot. All of this has been great learning. Unfortunately for early stage startups its hard to break in as a junior 😞 especially in this market where theres tons of ppl with more experience also looking for jobs

    • 0
      Profile picture
      Junior Machine Learning Engineer [OP]
      Taro Community
      a month ago

      If you or anyone else has any advice on how to tackle the breadth of interviews in this market I would really appreciate it.

      The nice thing about big tech is that how to prepare is largely a solved problem and is a function of time. Unfortunately right now big tech isnt hiring and early stage isnt willing to take risks on new grads/juniors so late stage startups/F500s are where im getting hits

A startup or start-up is a company or project undertaken by an entrepreneur to seek, develop, and validate a scalable business model.
Startups264 questions