From Business Student to Data Scientist in Tech
The First-Years Chronicles of a Data Scientist in Tech
Amongst the most frequently asked questions I often receive on LinkedIn, there is one that consistently stands out: How did I switch overnight from Business to Engineering and made it to become a Data Scientist?
So in this story, I will delve into my personal journey, and share with you the steps I followed, challenges I faced, and valuable lessons I gained that propelled me straight towards becoming a Data Scientist in Tech.
There is no one specific way to become a Data Scientist. As the saying goes, all roads lead to Rome. However, I am here to share one of the multiple ways this can be done. Especially for those starting with a Business degree and little to no scientific background.
The frontiers are blurry on the job market and knowing which type of Data Scientist you want to become will influence the skill set you need to have.
The most common archetypes of Data Scientists you will mostly find out there are : 1. Data/Decision Scientist: Leverages data to generate insights and value that will drive decision-making. Knowing Machine Learning is often required.
2. Research Data Scientist: Develops Machine Learning models as a product. Having a strong mathematics background is highly required. It is mostly a role occupied by phDs in Tech (but not only).
- Data Scientist/Machine Learning Engineer: A role at the intersection of a Decision and Research Data Scientist.
In my article, I will mostly be focusing on my journey to becoming a Data/Decision Scientist in Tech, which in my experience, seems to be the most predominant type of Data Scientist out there.
The Why
Before delving into the how, we first need to lay down the foundations for the why. Trust me, you will not go too far if you don't get your why straight and sound from the start.
The journey to become a Data Scientist is a challenging one, but surely one of the most rewarding too. As a matter of fact, the list of reasons why Data Scientists have one of the coolest and most coveted jobs today stretches long. For now, I will focus on the one that sits on top of my list.
To anyone asking me why Data Science? whether it be during interviews or simply curious people, my answer is always the same. I wanted to be a Detective, so I decided to become a Data Scientist.
You might wonder how these two have anything to do with each other? The first Data Scientists I've ever met instantly felt like they embodied what a modern-day Sherlock Holmes would be.
Data Scientists play with knowledge to solve puzzles every day. Ultimately, they spend most of their time investigating numbers to bring solutions to complex problems that only a keen analytical mind can solve. That is exactly what detectives do.
I have always felt a deep sense of excitement following the adventures of Sherlock Holmes. He __ restlessly pursued one clue after the other until the mystery was unraveled. I wanted to be an adventurer like him. But I hardly pictured myself dropping my studies to go solve crimes with the cops. I guess that felt a bit too extreme for my taste, and I was kind of hoping not to antagonize the bad guys this early in life. So being a Data Scientist felt like the best of both worlds.
It simply took some time for that epiphany to kick in. And by then, becoming one felt like a big stretch from what I was pursuing at that time – a Business degree.
Looking back, the only trait I shared with Sherlock Holmes was my knack for diving headfirst into seemingly impossible quests. They were far-fetched, considering I had zero experience with coding. But little did I know that embracing this leap of faith would be my golden ticket to pursuing my greatest passion in life: Music.
The How
Step 1 – Facing & Embracing the Math Monster
Despite a long-standing fascination for science, God had yet not deemed me deserving of the highly exclusive gift of deciphering scientific jargon easily. Nor was I blessed with perfect pitch either, but this has nothing to do with this.
So, anyway, upon graduation, I naturally gravitated toward what most clueless high school students that suck at scientific subjects do in this part of the world – Business studies.
My relationship with mathematics had been a rather tumultuous one. I was an economics major in high school. Physics left me feeling like I was in a parallel universe, and mathematical concepts sounded like secret codes from an alien civilization. Lessons took time to kick in, but.. they would eventually kick in sooner or later. Sometimes, very much later.
I knew very early on that carrying maths in my baggage would come in handy at some point. So I was not going to let my scientific shortcomings prevent me from dreaming beyond my abilities.
In my last year of high school, I decided to face the Math Monster once and for all. I invested all of my energy into taming the beast until it eventually yielded.
It was a game-changing moment.
It gave me the necessary confidence to later pursue a minor in Mathematics alongside my Business major during my university years.
Out of all the math courses I pursued during my minor, these are the fundamental ones that lay the groundwork for tackling Data Science and ML problems:
- Calculus
- Linear Algebra
- Statistics & Probabilistic Theory
For all maths-fearing souls out there, remember that we only hate something to the extent of how poorly we perform in it. The sooner we improve and excel in a subject, the more our perception changes.
So, if your goal is to become a badass Data Scientist, it's time to confront the Math Monster head-on and show him who's the boss!
Business was clearly not meant to be my calling. I enjoyed way too much getting tortured by the tantalizing clutches of the Math Monster. So in my final semester, I delved deep into exploring math-related career paths.
Ultimately my search led me to the field of Data Science. I seized an opportunity to intern as a Data Analyst and at the same time, got into NYU's Masters of Urban Informatics (a fancy word for Applied Data Science in the field of Smart Cities).
Step 2 – Building Intuition for Coding Takes Time
You may wonder how a Business grad with non-existent coding experience whatsoever managed to sway her way into an Engineering school? Well, remember those maths courses I took in undergrad?
It happens that you don't need much beyond those 3 courses to get yourself started with ML (Machine Learning).
NYU's degree was an immersive yet intense 12-month program where I was thrown right from the bat into ML modeling with Python, building databases with SQL, and wrangling Big Data on Spark (or at least trying to) all at once.
I have to be transparent here. The overnight switch from Business to Engineering was a hectic one. That year felt like an almost near-death experience all throughout.
Learning multiple programming languages for the first time meant cultivating a distinct intuition for each one, which takes time. Mastering all of them simultaneously, within a condensed timeframe, a heavy tuition bill at stake (and a pandemic at your doorstep) is not something I would advise for the faint of heart.
For someone who had never written a single line of code prior to this, this felt like a big shockwave to my brain that went full into overload. Poor thing didn't know how we'd gone from the principles of Management to running full-on ML models overnight.
In hindsight, one thing I would do differently would be to learn coding way before delving into Machine Learning, and not at the same time.
Playing around with ML was a blast, but that was far from being enough. I still failed to fully grasp what was going on behind the scenes of those Python packages.
You will never be considered a true Data Scientist on the Tech job market unless you are able to dig deeper. You must get fluent at explaining the underlying mechanics behind those prepackaged ML algorithms.
So I knew I had to understand those mechanics, but there was only so much I could achieve in one year coming from so far already.
At that time, COVID-19 kicked in and job opportunities in the U.S. were as non-existent as my prior coding skills. So I figured I could use an extra year in academia (or 2). I applied and got accepted into this unique dual program in France that combines the best of both worlds in Data Science: Business & Machine Learning.
Step 3 – From Library Importer to Explainer: Unleashing the True Data Scientist Within
Pursuing that degree turned out to be one of the best decisions of my life. It led me to study at the top Business & Engineering Schools in France, a feat that once seemed unimaginable.
I had never been the brightest cookie in school, but I have always had a talent in making my way through the roughest tracks. Being determined and stubborn were my most important assets, so I put them to good work early in life.
In the span of these two transformative years, I learned Data Science material I did not even know I needed to learn. The weapons I picked up along the way still guide me today and I am here to share them with you:
- Understanding the math behind ML algorithms separates the pros from the Python library users. I learned how to demonstrate mathematical proofs, but I believe grasping the key concepts behind the theories is sufficient. No need to meddle directly with the equations themselves.
- Being proficient with Python and SQL is an essential skill in the "Data Science in Tech" starter pack. Extracting and building pipelines for your data requires making a stop on data warehouses such as BigQuery, which usually are powered by SQL. While taking the time to master those Python basics will you help with data preparation & analysis.
- You will need Linear Algebra and Calculus to help you understand the fundamentals of ML theory, but nothing will beat having Statistics & Probabilistic concepts in your toolbox. Stats have direct applications in the daily work of a Data Scientist, so better make sure you understand the basics of statistical significance and probability distributions early on.
- Data insights are worthless if you can't communicate them to non-technical folks. Refining your storytelling skills is a continuous journey for Data Scientists, so better board the ship as early as you can.
- Cultivating resilience and patience when working alongside brilliant minds and experienced peers. Imposter syndrome can easily creep in, especially when you come from a different background. Self-doubt is a nasty beast, so better get acquainted with it early on to get rid of it.
Coming from far, it took me two years of training before I even dared call myself a Data Scientist.
My personal story is not meant to serve as an absolute guide on how to become a Data Scientist, as everybody's journey is unique. Rather, it is a testimony of one of the various ways this can be achieved. I hope my learnings inspire you in your turn to guide you in your own adventure.
By embracing these lessons, I've leveled up as a Data Scientist. They equipped me with the necessary tools I needed to kickstart my career in Tech.
The best programmers aren't necessarily the best Data Scientists. To thrive as a Data Scientist in the Tech world, coming from a business background, I learned that one must:
- Embrace the discomfort of mathematics. Master the basics of Linear Algebra, Calculus, and Statistics & Probability theory.
- Take the time to learn Python and SQL thoroughly. Avoid overwhelming yourself with too many programming frameworks.
- Dive into the mathematical foundations of Machine Learning to demystify prepackaged algorithms. Be ready to explain their mechanics.
- Hone your storytelling skills from the start. Master the art of communicating complex concepts in a compelling and accessible manner.
- Integrate business concepts with the knowledge gained from the previous steps to create a powerful combination.
In the past, I underestimated the worth of my Bachelor's degree in Business. It often seemed like I had spent years pursuing the wrong path.
But upon joining Spotify, I discovered that the fusion of Business and Data Science expertise creates the finest Data Scientists. The real power lies in those who can navigate both realms seamlessly. There is no reason why you should not be one of them!
