Introduction

Over the past few years, I have interviewed with multiple companies and have found success in landing the data science job, while also having tons of failures along the way. I made sure to keep note of what I was doing well and what I was not doing well during these interviews. Sometimes, I was lucky enough to get straightforward advice from the companies themselves. This article will lay out how to land the job as a data scientist, as well as some useful tips on the interview process.

Data Scientist

What is a data scientist?

Before focusing on how to interview for this position, it is integral in knowing the definition of a data scientist.

With my experience in several roles and also applying to countless companies, I have come up with my own definition. While it might not be 100% accurate for all positions, I think it serves as a valuable way to look at data science as a career.

• data science — automating processes with machine learning algorithms and other statistical methods with the use of programming languages like Python to ultimately solve a business problem.

Say you want to perform a linear regression or A/B test, this role might be more known as a statistician. Data science differs in that it requires coding languages and packages like sklearn and TensorFlow to automatically implement machine learning algorithms like random forest, for example. This algorithm would then be incorporated into a holistic data science pipeline that would:

— draw new training and evaluation data in with SQL

— clean that data and prepare it for the model

— apply a machine learning algorithm from a package

— evaluate new data and output predictions

— input predictions into a data table and/or user-interface

With a pretty inclusive and complicated process, data science can be overwhelming and there can be key points in the process that calls for specialization. This point is important in landing a job as it will determine what type of data scientist you will be. I believe there are two main types of data scientists.

While both roles have overlap, there are key differences in what each role focuses on. 

Business Data Scientist: 
- establishing the business problem
- collaborating with stakeholders
- explaining and interpreting model results
A data analyst or business analyst can also focus on establishing the business problem and coming up with key metrics or features that will ultimately be what trains the machine learning model.
Machine Learning Data Scientist:
- ingesting and querying data
- cleaning data and transforming
- coding the model
- deploying model into production
A machine learning engineer can also be a separate position that takes some of the roles from above and focuses even more on the last point - deploying the model into production.
A data engineer can focus on those first two points from above including ingesting and querying data, and sometimes cleaning and transforming the data.

#machine-learning #interview #towards-data-science #data-science #work #data analysis