Granular Data Scientist Interview Questions
& Process
Real candidates share what happened, how many rounds they had,
and how the experience turned out.
Based on 9 interview experiences · FREE TO READ
Candidate interview experiences
First-hand accounts from people who interviewed at Granular.
Granular
Data Scientist
Mid Average Negative experience No offer 4 rounds
Interview process
Take home Onsite
Interview formats
Technical Coding Other
Applied online and got a take-home assessment, then was invited for an onsite. The onsite has 3 interviews, each lasting an hour, where you're given a dataset and need to do classification. The first two are focused on timeseries modeling, so be quick. The last one is a basic SQL/Pandas interview. The Data Scientists there are typical SF 'coder bros', acting superior and quick to point out errors. It's tough to model a dataset without knowing the features and with the time pressure. The VP of DS does give feedback, but they should really re-evaluate their interview process. Honestly, I'd avoid this company.
Confirmed questions3 questions
- Model the data with location as a factor.
- Model the data with time as a factor.
- Basic SQL/Pandas questions.
Granular
Data Scientist
Mid Average Positive experience Decline offer 3 rounds
Interview process
Technical screen Onsite
Interview formats
Technical Coding Behavioral
The interview process here was enjoyable. The technical questions involved quickly analyzing and modeling toy problems. Since there wasn't much time for each assignment, practicing is super important. I thought the questions themselves were fun and engaging, and I liked working through the problems with the other data scientists during the onsite. The non-technical interviews were really nice, and I enjoyed talking with everyone there. Overall, I had a great interview experience and would definitely recommend interviewing here for a data scientist role.
Confirmed questions1 question
- Focus on rapid analysis and model development.
Granular
Data Scientist
Mid Difficult Negative experience No offer 1 round
Interview process
Take home
Interview formats
Technical
Agree with others. The take-home challenge is crazy, building models with two datasets for two problems in only 1 hour. I spent at least 10 mins just loading and exploring the data since there were no column definitions, so I had to guess what columns meant and if they could be used as predictors. 1 hour seems only enough time to fit really simple models like linear regression. My advice is to prep a R/Python template for linear regression (maybe with regularization) or decision tree beforehand. Once you get the data and problem, understand it fast, dump the training data into your model, and use something like cross-validation to tune and get your final model. Do it all super quick, even if it's a bit nonsensical - that's probably how startups work.
Confirmed questions1 question
- Can you build models using two datasets for two problems within 1 hour?
Granular Data Scientist Interview Questions
Quoted word for word from Granular interview reports.
“Can you implement 1-nearest neighbor?”
Read reports →“How would you predict missing values in a given dataset?”
Read reports →“How to build very simple models quickly?”
Read reports →“Can you build models using two datasets for two problems within 1 hour?”
Read reports →“Can you implement linear regression?”
Read report →“Given two datasets and two related exercises, perform inspection, modeling, prediction, and discussion within a one-hour timeframe.”
Read report →“What's your approach to handling missing values in a dataset?”
Read report →“Do you know some basic SQL-type queries?”
Read report →“Can you review the reports from previous candidates?”
Read report →Formats, difficulty and experience
Across all 9 Granular interview reports.