The structure of data science interviews varies according to the company and role, but most of them follow a typical format. The usual interview process is divided into three steps: a technical screening, a possible take-home assignment, and a series of onsite interviews.

The process is designed as a funnel for the company; they want to deploy their limited resources only on the best candidates. Here is how the funnel works:

• The technical screen exercise works to filter candidates who lack general data science knowledge.
• The take-home assignment is an assessment similar to the work that will be done in the role to test if your knowledge fits the specific needs of the company.
• The onsite interviews are usually carried out by your future colleagues. They evaluate each of the top candidates thoroughly for team fit and the data science skills required.

Each of these steps evaluates something unique about the candidate. Let’s break down the steps to determine how they work and see some tips to keep moving forward.

Technical Screen

The first round tries to evaluate your technical and soft skills as quickly as possible. Interviewers will continuously engage you with lots of technical questions trying to narrow down the list of candidates. You will need to manage tasks along the parameters of time and logic.

Questions are straightforward, so you’ll find them accessible if you learn the topics. The main challenge here is the small margin of error. If you make a mistake or don’t do well, you’ll most likely be cut before moving further along in the process.

Here are some tips to ensure you’re a cut above the rest.

1. Review a Little Bit of Everything

Technical screen interviews tend to come from fundamental concepts in data science, such as statistics, data design, data structures, and algorithms.

The questions tend to be quite direct because of time constraints. So if you understand the core data science topics, revision and practice should let you pass the screening without major issues.

2. Practice your Communication Skills

Communication is a vital part of the interview process. Interviewers need to understand your thought process and follow along with your solution because communication is also crucial for the job. As a data scientist, you’ll need to be able to share your processes and findings clearly, both to other data scientists and non-technical stakeholders.

Clear communication also shows confidence and helps reduce the tension in the interview. It lets our interviewers intervene to guide us if there is any misunderstanding. If we fail to communicate during the interview, we can incorporate incorrect assumptions into our final answers.

3. Ask Questions. A Lot Of Them.

You will face lots of ambiguous questions throughout this process. Seeking clarifications from the interviewer and reducing ambiguity will help you avoid misinterpreting tasks and ensure you’re aligned with their expectations. It will also show that you’re able to ask for better indications proactively when you’re lacking them. Interviewers expect you to ask for this clarification and speak to your communication abilities alongside technical knowledge.

Take-Home Assignments

After technical screenings, many companies offer you a take-home assignment. They do this to try to filter the best candidates, but not all companies include this step.

Take-home assignments tend to have a few defining challenges; here’s what you should know:

1. They Subject You to Extremely Ambiguous Requirements.

When doing a technical video/phone interview, you receive incremental feedback as you work on problems. Each interviewer has an idea of what they want, and if they see you going down a different/wrong/useless path then they’ll (typically) correct you. That feedback could range from providing guidance on assumptions to telling you the actual answer.

You get none of that in a take-home assignment. You are taking a non-standardized test every single time with a likely biased grader. You could spend the entirety of your effort analyzing data in an area the grader does not care about. We’ve seen assignments in which the requirements were as ambiguous as “analyze a dataset and turn it into a presentation” without any further clarification - not even when recruiters were pressed for details.

This is especially frustrating when you don’t move forward in the process since you get so little information on how to improve.

2. There Isn’t a Real Timeframe.

Interviewers usually give estimates in their instructions, such as “this assignment should generally take 3 to 6 hours” or “return it in around 2 to 7 days”. Most of us probably assume that every other candidate is actually putting in 6 to 12+ hours on the take-home assignment. And why wouldn’t they?

Take-home assignments grade your product against every other candidate who has also been given the assignment, so it’s direct competition. And doing a good job on the take-home means figuring out each edge case of the specific problem, plus the edge cases around the grader’s own biases.

Why wouldn’t you spend 12+ hours of your capturing these edge cases to gain an edge over your competitors? When reporting back to the recruiter, most applicants will say they were “able to finish it in like maybe three hours?”, so it can be hard to judge how honest you choose to be as well. In essence, there is every reason to understate the effort it took to create a really high-level product.

3. You Are Generally Given Zero Feedback.

What’s worse than spending those 12+ hours on an assignment and then not getting the job? An inability to determine where you went wrong and then use that knowledge to iterate on the next position that also requires an assignment.

4. You Are Telling Them That Your Time Is Worth Less Than Theirs.

Imagine interviewing for five different jobs that want coding assignments. Even if you only use the time estimated by interviewers, that is still 15 to 25 hours of take-home assignment drudgery over the course of a week or two… On top of your full-time job or job search!

This is all unpaid work. You can certainly say goodbye to the weekends.

There can be a cost-benefit analysis when facing multiple companies wanting take-homes completed; you will need to decide when and where the effort is worth it.

5. On the Plus Side…

On the plus side, take-home assignments can be useful for figuring out the work you would be doing on the job. Many times startups will take a sample of their data and thoughtfully give out assignments and questions that mimic actual projects for the role.

For those of you coming out of university, who often learned the material with datasets focused on education, take-home assignments will give you a good practical approach to data science using real-world data.

They are also a great way for unproven candidates to become competitive. Aspiring data scientists or graduate students should utilize the coding assignments and spend all of their efforts on making them perfect. It generally levels the interviewing playing field by allowing novice candidates that have more free time to demonstrate themselves through hard work and effort. And so, while take-homes are a detriment to some, they can be positive for others.

Onsite interviews

Onsite Interviews are the last step in the data science interview process. If you make it to this step, your future team members and colleagues will take the time to interview you.

There are generally several rounds, and each of them asks in-depth questions about a specific topic in data science. This is the part for which preparation gets so difficult due to how data science works.

Choosing the right interview questions to study for the onsite can be frustratingly challenging. You might religiously study machine learning concepts, only to get slapped in the face with a business case question. Or you’ll review a ton of case studies only to be asked about matrix multiplication in NumPy.

With this learning path, this won’t happen to you. In the next lesson, we’ll look at the different kinds of questions that appear in onsite interviews and how to prepare for them.