What kind of Data Scientist do you want to be?

Hey all, welcome back to my newsletter! Thank you for being here.

Today, I am talking about a topic that often confuses / surprises my clients…

Did you know there are different “tracks” of Data Science? And that depending on the “track”, you should have different skillsets and expect different types of interviews?

From my experience, there are two most common types of Data Scientists:

• Product Data Science (aka Product Analyst)

• Machine Learning Data Science (aka Research Scientist, Applied Scientist)

So what do these roles entail?

Let’s start with the common denominator between the two roles:

• Statistics — I’m talking about the fundamental Stats concepts here: probability, distributions and the basic ML models (regression, clustering and trees).

• Product understanding — this is about understanding the business and it’s products. Specifically, how do we use various data & analytics solutions to move the business forward.

• Strong communication skills — both verbal and written communication is important.

But what are the key differences between these two roles?

• Product DS focuses heavily on AB experiment design, analysis and interpretation to guide product decisions.

• Product DS is very business & strategy heavy, while Machine Learning DS is more focused & executional.

• Machine Learning DS involves deployment of models from end-to-end, while Product DS tend to use models to more deeply to answer a product / business question.

• Machine Learning DS tends to have a lot more heads-down time and long-term (3-6 month) projects, while Product DS tends to be more collaborative with short (1-2 week) projects.

So how do these differences materialize in the interview processes?

• Product DS interviews tend to have a LOT of product case study rounds. And it’s not just in your case study interviews, they want to see your product understanding across all interviews (including your stats, coding & behavioral interviews)

• Machine Learning DS interviews tend to have more coding & pure Statistics rounds. These mean including DS&A, system design, machine learning & deep learning tech interviews.

Real quick, if you’re prepping for DS interviews. Check out these ROLE-specific resources:

• Product Data Science Interview Guide

• Machine Learning Data Science Interview Guide

Thanks for reading until the end! See ya’ll in 2 weeks 👋