Glassdoor Data Scientist Interview Questions | Glassdoor.ca

Glassdoor Data Scientist Interview Questions

Interviews at Glassdoor

5 Interview Reviews

Experience

Experience
80%
0%
20%

Getting an Interview

Getting an Interview
80%
20%

Difficulty

3.0
Average

Difficulty

Hard
Average
Easy

Helpful (4)  

Data Scientist Interview

Anonymous Employee in Mill Valley, CA (US)
Accepted Offer
Positive Experience
Average Interview

Application

I applied online. The process took 6 weeks. I interviewed at Glassdoor (Mill Valley, CA (US)) in November-2016.

Interview

I applied on the Glassdoor website, got a response from a recruiter pretty quickly. Had a follow-up phone screen with the recruiter and 2 with members of the team (1 informational, 1 general case phone screen). Next, there was a take-home data assignment. Last step was on-site at the Mill Valley office with 5 interviews going over past machine learning projects, understanding metrics, business communication, product intuition, simple coding knowledge, a few brain teasers, etc.

Overall, the process was pretty straightforward and positive with a lot of communication from recruiting and from the team.

Interview Questions

  • What are the advantages/disadvantages of different metrics for evaluating search results? (Metrics were provided and defined)   Answer Question

Other Interview Reviews for Glassdoor

  1.  

    Data Scientist Interview

    Anonymous Employee in San Francisco, CA (US)
    Accepted Offer
    Positive Experience
    Average Interview

    Application

    I applied online. The process took 6 weeks. I interviewed at Glassdoor (San Francisco, CA (US)) in May-2018.

    Interview

    I applied online in March, 2018. The whole process took 6 weeks including two technical phone interviews and an one-day on-site interview consisting of four individual interviews.

    The first phone interview is about assessing my coding skill as well as my experience. Pretty standard coding question that is similar to questions in Leetcode.

    The second phone interview is about my experience and some machine learning design questions. This interview happened two weeks after the first phone interview.

    Finally, the on-site interview is a whole day interview starting with a lunch with prospective team members. After that, there are four individual interviews with different engineers/scientists with different types of questions: background checking about my knowledge in Machine Leanring, probability questions, describing past projects, coding questions, machine learning design questions, my motivations when joining Glassdoor. After the fourth interview, I met the recruiter again and she told me that Glassdoor would like to give me an offer. The day ended with having dinner with my prospective manager.

    Overall, it has been a nice experience for me to have interviews with Glassdoor. I do feel a warm welcome from Glassdoor during my on-site interview.

    Interview Questions

    • Q: Write a function to do integer division with strings as input
      Q: Why do you choose Glassdoor?   Answer Question

  2. Helpful (2)  

    Data Scientist Interview

    Anonymous Interview Candidate
    No Offer
    Negative Experience
    Average Interview

    Application

    I applied online. I interviewed at Glassdoor in April-2018.

    Interview

    Initial phone screen followed by technical phone interview with hiring manager.

    I think Glassdoor get the prize for worst data scientist interview question ever. It starts out fine, the hiring manager goes through your machine learning projects from work and talks about the technical details of its implementation. Finally you get to the coding portion of the interview. And you are asked to convert strings to binary algorithmically. Now if you are a computer scientist/CS graduate you may remember this. However, many data scientists do not come from a CS backgrounds, and I am pretty sure the hiring manager knows this by looking at your resume and actually interviewing you. It's odd because of the dissonance between the actual real world work and experience that is necessary to do data science work and the absolute irrelevance of the question. And it is not even one of the more common irrelevant algorithms/data structures coding questions that are asked during data science interviews, e.g. merge sort.

    It is really strange. One can have the relevant data science skills and experience, but be cut down by such a question. At the very least if they are going to ask an algorithms question it would be something like regression, why an interviewer would choose something like string-to-binary over something that is relevant to the job is hard to understand. However if you frame this issue in the context of bias and elitism then you may have an explanation.

    Interview Questions

    Glassdoor Response

    9 Nov, 2018 – SVP Engineering, CTO

    I'm sorry you had a bad experience. In our view, this question doesn't require a great deal of Computer Science. The common solution uses division and remainders...and then standard coding constructs... More

  3.  

    Data Scientist Interview

    Anonymous Interview Candidate in Mill Valley, CA (US)
    No Offer
    Positive Experience
    Average Interview

    Application

    I applied through an employee referral. The process took 3 weeks. I interviewed at Glassdoor (Mill Valley, CA (US)) in September-2018.

    Interview

    Had an initial call from a recruiter followed by a 30 minute screen by a hiring manager that included a brief overview of the position. The next step was a take-home data analysis assignment that took about 3-4 hours to complete.

    Afterward I was called into a 4 hour on-site interview where I spoke with 5 different people about my experience and how I'd approach example issues Glassdoor had along with a couple basic questions about probability.

    In all the interviewers were friendly and respectful, but unfortunately the position was not a match.

    Interview Questions

    • What sorts of exhibits would you show a product manager to help them understand what parts of a self-service portal employers might be having trouble using.   Answer Question

  4. Helpful (8)  

    Data Scientist Interview

    Anonymous Interview Candidate
    Declined Offer
    Positive Experience
    Average Interview

    Application

    I applied online. The process took a week. I interviewed at Glassdoor in May-2015.

    Interview

    -Received a dataset to analyze at home. Got the week-end but spend about 3 hours. This "homework" was interesting (unlike some other completely useless homework that I got).

    -Skype interview with a data scientist, classic background question and some question about the homework. Not very technical, but interesting. They proposed interesting projects, but got a better opportunity somewhere else and refuse the offer.

    Interview Questions

    • - Difference between classical regression/classification methods. Mainly tree based methods (boosting vs bagging vs RF).
      - Cool projects, cool experiences in the field of data analysis
      - What do people say about me   Answer Question

Don't Miss Out On a Job You Love
Upload a CV to easily apply to jobs from anywhere. It's simple to set up.