Data Scientist? Here’s How to Ace Your Next Interview

Expert Advice to Avoid Rejection Letters

Have you ever interviewed for a data scientist job and received the following communication in your email inbox?

“We regret to inform you…”

“Thank you for your interest in our company and for interviewing. We have decided…”

The chances are that you’ve probably received a communication or two like this throughout your career. You spent time searching for the right job, researching the right companies, filling out long applications, acing the phone interview, and then after the in-person interview you didn’t get the call back.

“What went wrong?” you may wonder.

If you were lucky, the hiring manager may have given you some feedback. But for many data scientists, a rejection is a quick email or phone call stating that you simply did not get the position.

We asked a few hiring managers for their suggestions about what typically goes wrong at the interview to prevent a candidate being taken forward to the next stage. From our discussions, we’ve compiled the best tips to help you rebound after the initial rejection.

Seek to Improve & Grow

It’s safe to assume that all candidates who are invited to an in-person interview for a Data Scientist position have similar basic skill sets and qualifications. How can you stand out from your competition?

According to Dr. John Aven, PhD, Director of Engineering, “Most bootcamp-level data scientists have some major shortcomings. The first, and foremost, is that they neither have a deep understanding of the mathematics they are applying nor the ability to say why they are using a certain algorithm.”

Dr. Jack Wang, Lead Algorithm and Machine Learning Engineer adds these pearls of wisdom for data scientists:

“1. Practice, practice, practice. Knowing the theory behind data science is well and good, but usually I find that people with little practical experience have a harder time finding ideal data science positions. One resource I recommend is kaggle.com to get data sets and practice.

2. Constantly look for new things to learn. Data science/machine learning is an extremely active field. New things come out on a daily basis, it is important that a data scientist keeps up with the latest development in the field.

3. Have a strong engineering background. Data scientists often ignore foundational computer science/engineering concepts, because they believe that they work with data and statistics, therefore engineering is not/should not be part of their toolbox. That thought is very dangerous. In fact, every company that I’ve talked with had a requirement of at least some understanding and experience of engineering, writing code that performs and having the ability to take the code all the way to release in production are very important.”

Some of Aven’s top tips for Data Scientists to upgrade their own skills include reading books to improve both soft skills and technical, math, and data science skills. He suggests:

·      To improve your soft skills (because Data Scientists must be consultants)

·      To improve technical, math, and data science skills

Dr Aven also recommends that you read anything by Andriy Burkov and “anything that gets you down to basic and has you build from scratch.

In addition to suggesting reading material, Dr Aven says, “Get involved with online competitions (e.g. Kaggle and related tech skills), learn Python OOP (there is lots of stuff out there), and try getting certifications (even though they are just paper) from the Python Software Foundation (Coursera and similar don’t mean anything).

“Learn a cloud platform and get certified (and build stuff there really) – just choose one and go – but learn basics of what you are doing before you do something foolish. Learn Docker and Kubernetes – there is a lot of new stuff happening there (Kubeflow, Kubeless, Knative, Polyaxon, Seldon, and Argo). Learn Dask and Ray, and get experience with Rapids.ai. Find new things and experience them.”

Make Sure You Stand Out and Offer the Most Value

At the end of the day, making the right hiring decision isn’t always about how much the hiring manager likes you. Every year companies have a budget they are willing to spend on new talent. Your challenge is to show them where you fit in. Director of Analytics, Dr. HJ Wassenaar suggests that it often boils down to a few basic and straightforward steps.

“In my experience as hiring manager,” he says, “the candidate often doesn’t have sufficient problem-solving skills. That can be helped by knowing lots of different statistical and modeling techniques – developing a deep toolbox. Also, especially important is the understanding of the basics of probability and statistics. It’s the latter where many of my candidates stumble.”

“If you know that you have an area of weakness that your competitors may have as a strength, it could be time to sharpen those skillsets. It’s definitely beneficial to make sure that you are able to understand the work at a deeper level.”

On that note, it’s not always your skillset that sets you apart. Dr. John Aven, PhD adds, “Without soft-skills a data scientist is maybe only slightly more than an academic statistician. If they cannot integrate with the business, talk with them to discover needs and discuss solutions, and so on, then the solutions they create will be manifestations of their own imagination and have zero business value. This is an issue I have seen happen numerous times – and is a big driver as to why you hear stats like 7/10 data science projects never make it to production.”

Ask yourself some key questions such as:

  • Are you taking your education, projects, knowledge, and skills and applying them in a way that is valuable for the business or company?
  • What contribution have your past projects made to the company in terms of real measurable results?

Being able to demonstrate positive answers to such questions during an in-person interview will help to show that you are an employee who adds true value to the teams you work with.

Stop Doing These Things

Sometimes it’s not what you didn’t do, but something you did that prompted the rejection. It could be what you said, how you said it, or something you studied that seems off topic. Let’s dig into a few common mistakes.

  • Stop learning new skills that are irrelevant or add no value to the work

For example, some hiring managers don’t care for Coursera or bootcamps. It’s less what you studied and more ‘what value does it add?’.

Can you show that the certificate or course you took makes a difference in your quality of work? Is it something that saves time or money? Focus more on demonstrating how you apply your knowledge, not just checking a box that you have it.

According to Dr. Wassenaar, “Paid programs like Coursera help but they are not a must. If you know your stuff, you know your stuff. You don’t need an expensive course for that.

Dr. Aven also stresses the need for the candidate to understand concepts at a deeper level. “You can’t get away with just saying you took a class. Be prepared to prove you fully understand that knowledge.”

·      Stop relying solely on your education

Dr. Wassenaar states that when he is looking for a data scientist he focuses less on the knowledge and more on experience.

What can the candidate do with that knowledge? That’s what I’m interested in. Having relevant work experience on your resume helps to get past the initial screen. If they are lacking experience, the candidate could list a relevant school or personal data science project.

In the interview, if all you have to show is a certificate or degree, you must find ways to highlight personal projects or work experience that showcase you don’t just talk the talk but you can walk the walk too.

·      Stop making your skills so niche – focus on your contribution to the big picture


Remembering that companies have a hiring budget and can only hire so many people to their teams in a given quarter or year, it’s imperative that you bring as much to the table as you can. However, you want to make sure it’s not any old skill, but the right skillset.

Therefore, you should seek to broaden your knowledge in ways that make you useful to the company. When asked about hiring Data Scientists, Dr. Aven stated:

“Someone with machine learning and engineering skills (or a Data Engineer with machine learning background) will offer greater cross-functionality. They will write better code, and, in doing so, deliver a better product that will get into production faster, present fewer errors, and is more easily maintained by others.

“However, adding DevOps skills is important. This means that the Data Scientist can write unit tests against their code, build continuous integration, continuous delivery, continuous testing, and work the entire data science lifecycle.

“The continuous testing part is unique. It is the MLOps component that differentiates data science from traditional software engineering needs. It’s a huge bonus if a Data Scientist can manage infrastructure as code as well.”

Dr Aven’s logic here is easy to follow, and shows how such a grouping of skills makes a Data Engineer indispensable to the company.

Summing Up

Losing out on a great job opportunity is difficult for anyone, and it can be hard to recover from. The best thing you can possibly do is to find out what went wrong, look for ways to improve, make sure you stand out, and streamline your education, skills, and experience so they scream the value you add to the team. If you implement these tips, then you may be able to open doors you only dreamed of before.

Need help making these changes? Need help starting a passive job search? Reach out to Kofi Group today and let our executive recruiters give you their guidance and coaching that will help you land the job of your dreams.