Not “Mission Impossible”: 6 Lessons Learned from my Journey of Becoming a Data Scientist

Published in

Wix Engineering

5 min readApr 13, 2023

Learning mode ON

Transitioning from a Data Analyst to a Data Scientist can be challenging, but many experienced analysts aspire to do so. To make a successful transition, it’s important to step out of your comfort zone and approach the new role with a learning mindset.

If you’ve decided to take the risk and make the transition, where should you start, and how can you do it the right way? I’ll answer these questions and provide you with a plan, but before doing so — let me introduce myself.

I’m Omer Dodi. I was a Data Analyst at Wix.com for two years and recently transitioned as a data scientist in the Cyber Security domain.

I have the privilege of working at a data-centric company with a leading AI group and a very supportive manager, which provided me with the foundation to start to think about this transition.

As an analyst, I was able to work with data scientists on many different projects and understood from my experience working with them, that I wanted to become one of them.

But, to achieve this goal, I needed to learn new skills.
And it sounded like A LOT of new skills.

So, where did I began?

Let’s deep dive into it!

Skills you have VS. Skills you need

Data Analysts and Data Scientists do share some comparable skills.

As a Data Analyst looking to transition to a career in Data Science, it’s important to identify your existing skills and assess how they can be leveraged in your new role as a Data Scientist.

Let’s explore some of the essential skills that are critical for a successful transition.

50% of a Data Science Project Process is Data Analysis Work.

This is not accurate and not true for all projects, but it made my initial point of view much calmer for sure.

Data Science projects typically involve analysis tasks before and after the Modeling and Production stages.

Each project starts with a problem definition, followed by data collection, data cleaning, data processing or exploratory data analysis. And at the end of each project, there’s a data evaluation process and some contain data visualizations.

All of these are analysis tasks. An average analyst could learn these, but will have to learn a few basic skills as a starting point.

Talk fluent Python

The most basic preparation you can do before starting to learn Data Science is to learn how to code using Python.

Learning the basics of Python, as well as key data analysis/science libraries such as NumPy, Pandas, Matplotlib, Scikit-learn, and Seaborn, is essential.

Python has many free online learning resources. A great one to start with is Tech with Tim (YouTube channel or his website).

Make sure to understand Probability Theory and Statistics

Before starting to learn Data Science modeling methods make sure you’re familiar with (or remember) all relevant statistical terms in Probability Theory.

There are many books with basic Probability Theory, such as “Grinstead and Snell’s Introduction to Probability”, a super long book but I recommend just browsing through it to make sure you’re familiar with the different subjects.

Enhance your MLOps toolkit

While learning Data Science, a major focus will be on Modeling in theory and practice. However, there are additional “must-know” concepts that may not be covered in depth, including:

Git & Github
Cloud Hosting (choose one, I chose AWS)
Shell (choose one, I chose Bash)

I find that these three topics are crucial for my day-to-day work as a Data Scientist, as they can help me solve problems and gain a deeper understanding of the project process from start to finish.

There are many online resources available for each of these topics.
Here are few examples:

Also, I would suggest developing a basic understanding of computing power in order to accurately calculate run time and memory usage of Python code. So by just learning hardware basics, you can position yourself for success in the field of Data Science and stay ahead of the curve in an ever-evolving industry.

Start a Data Science program

After gaining a strong foundation in Python, probability and additional concepts, I decided to enroll in a 1-year Data Science program at the Y-Data School of Data Science.

While there are other programs available, I can personally attest to the quality of this program and express my gratitude to the awesome managers and lecturers of the program.

Not a Mission Impossible

Becoming a successful Data Scientist requires dedication, hard work, and a willingness to continuously learn.

By mapping out the subjects discussed above, one can start and plan for transitioning to a career in Data Science and understand the necessary effort and commitment required to accomplish it.

And for an optimistic closer, here’s an interesting Q&A on the subject from a post in Quora.com: