How to prepare for a Data Science interview

Welcome to the Data for Everyone newsletter! In this week's entry I'll discuss what is like to interview for Data roles, it's difficulty and some tips & tactics that you should deploy in order to stand out and land your desired role.

I know that I might sound like a broken record at this point but there's never been a better time to launch your career in Data Science. The median salary of roles in the field exceed the 100k and the opportunity to grow is much bigger than any other fields in tech like software engineering.

And though everything sounds like rainbows & sunshines... In order to get to the position first you have to actually demonstrate that you can get the job done, and you'd be surprised but the process to actually do that is not bread & butter.

Prepping for a Data Science interview is not like preparing for other job interviews. It's even more similar to other tech roles, including multiple rounds of interviews (Getting up to 7 rounds of interviews in some cases, most of them being hard-coding interviews) and with a high-level of complexity.

Where to start?

Research the role & company

Probably the most underlooked aspect of the job interviewing process. Companies & recruiters specially are looking to see that a candidate takes the time to understand the company, what it does and what the specific data role would bring to the company in question.

So, first things first, research the company's website. Look for the company's basic information like it's industry, it's mission, it's trajectory for the last couple of years...

Follow the LinkedIn page so that you're also updated with the company's latest information regarding job postings, social presence and it's labor force network. Try to connect with employees in your desired role so that you can ask for feedback regarding the interview process as well as the role's requirements and what it looks like on a daily-basis.

Tweak your portfolio & latest projects

If you read last week's blog entry, you're now aware of the importance of a good portfolio site. If you haven't, check last week's entry.

While you won't get hired on your portfolio alone, your data science portfolio could be the key to land the interview. Is because of this that you should spend as much time on your portfolio as you would in resume (If not more).

Tweak your projects so that they adjust to the job requirements, even create projects related to the company itself. If you're applying to a social media company create a project analyzing your own instagram data (Or your LinkedIn's). Be smart & show the company that you have the iniative and the skills beforehand.

Review theoretical concepts

In order to ace your interview, you will have to demonstrate your technical accumen, no matter the level you'll be applying to, recruiters want to see you know the basics. Some technical concepts that you should review before any data interview are: Probability, stats, hypothesis testing, A/B testing and more which I'll be discussing in further entries.

Practice your technical skills

A few technical skills that you should be able to demonstrate in your interview happen to be the following (usually):

Statistical Analysis

Most data science concepts are a computational expression of mathematical ideas. This is why you need to have a solid understanding of math and statistical concepts and apply it through different programming languages.

Make sure that you have a good understanding of variability, probability distributions, logistic regression, linear regression, and statistical significance... And how to perform them in Excel, R & Python.

Data manipulation, extraction & analysis

All data professionals need to be good at working with data... but also need to understand the whole process.

This goes from data selection and extraction of the data you want for analysis. The data preprocessing and cleaning (Where you can also show off your skills of data manipulation) to get to the main stage: Data analysis.

I know this sounds like a lot, so the best thing that you can do is start practicing by working with datasets that you can find in public sources like Kaggle.

Hard-coding

Though not all data employees have a programming background, it's very useful to understand programming languages and techniques. Python & R are the two choices to pick from when getting in this field.

Hop into data courses, watch YouTube videos, analyze your own datasets and get more comfortable with the IDLES until you get to the point of fluency programming.

Modeling and data visualization

Presenting your findings is the most import part of a data job, even more when you're in the middle of a job interview. You could have the best data analysis skills but If you can't convey your skills and present them so that your audience understands the insights you found... You gotta practice more then.

To do this, translate your findings into charts and visualizations. You can do this through softwares like Tableau or Power BI.

To practice this aspect I suggest that you create some sort of visual for all of your portfolio projects. This way you can get more comfortable & creative when showing results.

Now, what can you expect from a Data Science interview?

Online Tests

Between an hour and two, this is an online screening normally filled with behavioral questions and some theoretical concepts. Just a stage to filter out candidates.

Phone Interview

Once you pass the first stage, a recruiter will set up a call with you that will last from 15 mins to half an hour. This is a general conversation to get to know you. On this stage you will be asked the typical questions like tell me about yourself, career plans, salary expectations...

Technical Interview(s)

The most important part of the recruiting process. Is during this round(s) that you will be tested on your core data skills and knowledge.

This can be done in many different ways. From coding challenges that you would have to complete in an hour, cases that you will have a week to complete generally involving the previous stages mentioned followed by a presentation or an hour-long technical interview asking you about concepts.

There's many combinations on this stage, which happens to be the most intense and challenging out of the whole recruiting process. Hence my emphasis on you preparing well for this stage.

HR Interview

Once you're done with the technical part you'll be tested on your soft skills as well as your communication. At this point the company is ensuring that you're a good culture fit for the company.

Leadership Team Interview

Though uncommon, sometimes you're asked to meet with part of the leadership team so that they can make sure that you'll interact well with other team members and that your cooperation skills are on point too.

And this is the recruiting process for you to land your first data job. If you're interested on resources to ace the different stages of the recruiting process I suggest you check the following resources I've prepared for you to Ace your Data Science interview (And yes, they're free).

I hope that you enjoyed this week's entry! If you liked the article make sure to connect with me on LinkedIn to stay updated with my newsletter, see you next week!

-Alfredo S.