How to interview entry level Data Scientists and interns

25th January 2019

I work in the Artificial Intelligence Center of Excellence at Pfizer, a global pharmaceutical company. As is common with larger companies, we often have a “trial” period for many incoming employees who are just starting out. I myself was an intern at Pfizer for a bit before I became full-time.

It’s always interesting having sat on both sides of the table — obviously for some time I was a data scientist looking for a job. Now I’m a data scientist hiring other data scientists. Having so recently been in your shoes, I want to provide some insight on how I interview and what I look for in my data science interviews.


Like anything, these things probably aren’t universally true; every job is different. As Obi-wan Kenobi said: Only a sith deals in absolutes. Still, I think they’re all important and could potentially help you out in any entry-level data science interview you take.


Sorry, this is pretty important. Regardless of how you do in the rest of the data science interview this is like stage one — the bare minimum requirement before moving on to general and personal questions. While I won’t divulge the exact questions, here are the things I look for:

Programming (namely Python) knowledge

Image result for python code gif

I always liable to ask a question on general programming knowledge. Being able to code a simple algorithm is a necessity — and many more bonus points for making it efficient. Regardless of your other qualifications, your programming knowledge will showcase your ability to actually generate tangible value to the team.

Machine learning — or lack thereof

If you’re a data scientist, you need to know what problems you should apply machine learning for — and more important when you shouldn’t. 

Image result for gradient descent gif

Asking your thought process on some potential business use cases is a good way to understand how you approach challenges. Some cases might be simple solves: “I’d use this algorithm because this”. However, a data scientist also needs to remember that machine learning is not the solution to all problems. There are a number of tasks that can be solved with analytics, automation, or simple tweaks to an upstream part of the process.

Miscellaneous CS and stats

Outside of Python and machine learning there’re a number of other technical things necessary for a successful data scientist. Linux, Spark, or a basic handle on descriptive statistics and probability. 

I don’t necessarily ask these in my data science interviews, but someone on my team almost surely will. It’s good to have a handle on the data science technical landscape, and research what common things may come up during your data science interview.



While your technical answers can easily determine whether or not you’re a good hire, these aspects are where you can really shine. For me, some of them are arguably more important (though please don’t forsake technical study because of that). 

Problem solving

First and foremost, the data scientists on our team are problem solvers. We come up with novel solutions to business problems using a variety of different techniques both in and out of the AI/ML space that require creativity and critical thinking.

This could involve brain teasers, general questions, or just an assessment of your problem-solving experience. But I want to see if you have the hackathon-oriented mindset necessary to deliver value to our team. 

Creativity can be an underrated aspect in a data scientist, and I’m sure at some companies it isn’t a trait that provides much value. However, if you’re looking to be a problem-solver and not a boxed-up data scientist assigned to a specific task, you’ll need to harness creativity and critical thinking effectively.

Interest and Willingness to Learn

When I say interest, I mean both in the specific job and in the field. Another hallmark of data science is how fast the field is evolving. If you’re just in it for a salary boost or because “AI” looks good on a resume, it can be easy to sniff out.

What I really want to see is a drive and ability to learn. For me, this is the most important quality you can possess. If you are genuinely interested in the field and convey that enthusiasm, then I can be sure you’ll apply yourself wholeheartedly to improving both yourself and the team’s technical toolkit. 

If you have your own personal data projects, take courses in your spare time, or are just generally interested in solving problems with data then you are worth the investment. You will not only accomplish what we need right out the gate, but will always be bettering yourself for the sake of the team or the goal. Driving innovation is a huge part of what we do, and we need you to stay up to date with technology and technique to contribute to that.


Personally, I don’t like to pull from a list of questions. Instead, I like to hold a general or technical conversation and just see where it goes. How do you convey complex technical explanations? How will you explain model results to a marketer? A salesman? These are critical post-technical aspects that can make or break the success of your team. 

Beyond that, your data science interview is about how you’ll fit in with a team. Data science has a number of key aspects, but what doesn’t change from normal interviews is you’re just trying to make people like you. Being charismatic never hurts your chances . How you get along with others can really provide business value when it comes to group coding, meetings or just workplace banter. 

And that’s it!

Thanks for reading! Feel free to reach out to me on LinkedIn and good luck in succeeding on your data science interview and getting that job!


Read also: What to watch in 2019 as an HR professional

Tagged with:

Kyle Gallatin

A data scientist with an MS in molecular & biology. He currently aids to deploy AI and technology solutions within Pfizer’s Artificial Intelligence Center of Excellence using Python and other computer science frameworks

To give you the best possible experience, this site uses cookies. Continuing to use means you agree on our use of cookies.