How I Became a Data Scientist at Meta — No Stats Degree; No Bootcamp

My 6 jobs and 2 career pivots

Mandy Liu

Aug 01, 2024

Photo by Muhammad Asyfaul on Unsplash

Ever dream of transitioning into data science but worry you don’t have the right background?

Maybe you think you are behind the competition without a traditional degree?

Or you aspire to work at a big tech company but don’t think you are worthy?

If any of these thoughts crossed your mind, you're not alone.

I was in your shoes. And I’ve spoken with many aspiring data scientists facing the same doubts.

This article is for you—those who want to break into data science without the typical credentials. While it may not answer every question, I hope it gives you clarity and confidence to pursue your dream career.

Here’s a glimpse into my journey, broken down into six phases:

Discovery
Desire
Investment
First pivot
Second pivot
Opportunity knocks

🕰️ Reading time: 11 minutes

My Journey in a Nutshell

2016: Graduated with a B.S. in Economics and an M.S. in Commerce.
2016: Began my career as a consultant at Ernst & Young (E&Y).
2018: Transitioned to a growth analyst role at a dating app in Beijing.
2018: Moved to London, continuing with E&Y.
2019: Pivoted to a data science consultant position at a SaaS startup.
2020: Landed an offer from Facebook (now Meta) as a data scientist—a dream come true!
2023: Joined Nextdoor as a marketing data scientist.
2024: Launched my solopreneur journey, starting a weekly newsletter,
The Curious Techie
, to help others land jobs, grow their careers, and develop new thinking frameworks.

Now that you know a bit about my path, let’s dive into each phase and explore how you can carve your path in data science.

1. Discovery: gimme more of that data thing

graphs of performance analytics on a laptop screen — Photo by Luke Chesser on Unsplash

When I started college in the U.S., I chose to study Economics—a field that combines numbers and logic, which I love.

Concepts like externalities, supply and demand, and opportunity costs fascinated me.

There was only one problem - it was bloody impossible for a visa-begging international student to find a job sponser, especially with a non-STEM degree.

So I decided to stay in school for another year to pursue an M.S. in Commerce at University of Virginia (UVA).

Though I chose the program because it was top of the country, if I was honest with myself, I wanted to stay close to my boyfriend at the time who was joining the program.

Cute? I know. Not my best decision—something I detail as regret #4 in my 9 Regrets as a 30 something article.

During this one-year program, I specialized in Business Analytics. This was arguable the closest I’d been to data science acaedmically, yet it was still miles away.

While the program introduced me to data science concepts - programming languages, relational databases, and machine learning models - it focused on business cases and commercial acumen.

It left me wanting more.

It left me wanting more coding, more visualization techniques, and bigger datasets.

As recruiting season approached, my primary focus was finding a job to stay in the U.S. I was desparate for any job to buy me a ticket to board that American dream.

Luckily, I landed my first role as a technology consultant at Ernst & Young (E&Y) in New York.

The job was 15% technology (mostly Excel and PowerPoint) and 85% financial regulations—a world I found dull.

The long hours and frequent travel wore me out quickly, making it clear that this wasn’t the career for me.

When my project in Charlotte ended, I sought out more technical roles and found one in reporting, using Tableau and SAS. I didn’t know a single dime of SAS so I worked my a$$ off to learn on the job.

This project laid the technical groundwork for my future in data science and ascertained my belief: I like working with data and I want to get more technical.

Takeaways:

Act on your interests: Once you have an inkling of what you enjoy, pursue it to test your hypothesis. It’s a cost-effective way to explore your interests.
Seek relevant opportunities: If your current role doesn’t align with your career goals, look for projects that do. Don’t hesitate to take risks and learn something new on the job.

2. Desire: craving a cool office and interesting problems

Picture this: It’s New York, and I have a close friend—let’s call him C—who works as a data scientist at a startup.

Every time we hung out, he’d rave about how awesome his job was. I’ll admit it, I was envious. He got to wear jeans and sneakers to work while solving fascinating problems like optimizing where and when to show pop-up ads. It sounded so cool.

Then, one day, I visited his office. Imagine a bright, open-plan space filled with ping pong tables, lush green plants, a relaxed atmosphere, and best of all, free food and snacks.

a room with a table, chairs, and potted trees — Photo by Bernd 📷 Dittrich on Unsplash

It was my first glimpse into the world of tech offices, and it blew me away.

Sure, later on, after visiting places like Google and Meta, I realized his office was just the child’s play. But at that moment, I knew: I wanted what he had.

Seeing firsthand the life of a data scientist was the spark that ignited my desire to dive into this field.

3. Investment: hustle, hustle, hustle my a$$ off

Eager to dive headfirst into a new field without formal training, I practically begged C to mentor me in data science, and thankfully, he agreed.

This was a game-changer for me.

Learning Python: a revolution that saved my career

C convinced me to invest in Python instead of going for a SAS certification, which I had considered since I was already familiar with SAS at work. But he emphasized that Python was not just the future of data science—it was revolutionizing many fields. He introduced me to essential Python libraries like pandas, numpy, and seaborn.

I was hooked immediately. Coding in Python was electrifying, like discovering a magic spellbool - it felt like cracking the cheatcode of data analysis.

While my early attempts were clunky, C encouraged me by saying my code demonstrated “innovative ways to solve problems.”

The stats monster: facing the necessary evil

Next, C insisted I tackle An Introduction to Statistical Learning, the textbook from his Stats Masters at Columbia.

Reading a textbook isn’t exactly a thrill ride, but I knew this was a necessary step. And yes, while I promised no stats degree in the title, let’s be real—you need a solid grounding in statistics.

This book is a beast, covering everything from fundamental statistical models (regression, classification, trees) to unsupervised learning, complete with exercises in R and Python. I dug in, page by page, exercise by exercise, trusting that Columbia wouldn’t steer me wrong.

Getting more hands-on: the power of personal projects

To top it off, C helped me find a personal project that would deepen my Python skills and keep me engaged. I learned web scraping with Beautiful Soup, a popular Python library.

It was like unlocking a new level in a video game—I realized just how powerful Python could be.

Takeaways:

Find a mentor: Someone who not only teaches you but also keeps you accountable. If you can’t find one near you, consider going to network events or attending webinars, or look for one on on sites like topmate.
Read key texts: An Introduction to Statistical Learning is a must read, at least the first half.
Avoid tutorial hell: Get your hands dirty with coding projects.
Pursue passion projects: Find something that genuinely interests you to keep the learning process engaging.

4. First pivot: a bold move to Beijing

Imagine ending 2018 with your life in limbo.

That was me—packing my bags and leaving New York after 14 months because I missed out on the U.S. work visa lottery.

To top it off, my transfer to E&Y London hit a snag. Facing an immense shortage of medical professionals, The UK prioritized visas for doctors and nurses, leaving my application in a bureaucratic black hole.

So, I made a bold move: I headed to Beijing, the heart of China’s booming tech scene, where over 100 companies hit $1bn valuations in 2018.

The tech sector was in a massive boom. Companies hired lightening fast, valuing your potential over just your credentials. In one week, I snagged four job offers. I must’ve convinced someone I wasn’t entirely clueless!

I chose to join Tantan, a leading dating app, as a growth analyst. It was an easy decision—exciting challenges and free meals? Sign me up!

Me at the Tantan office. Image by author

When I started, my SQL and Python skills were basic at best, but my eagerness to learn was off the charts.

Within seven months, I sharpened my SQL skills, built a basic churn prediction model in Python, and collaborated with PMs and engineers on AB tests.

My confidence soared as I realized that I could indeed pursue a career in tech and become a “data wizard.”

After 10 months of anxious waiting, 70+ email exchanges, and several expensive international calls, my UK visa finally came through. I transferred to London with E&Y, but I knew my journey in data science was just beginning.

Takeaways

Target digh-demand, low-supply markets: Think beyond traditional tech hubs. While Beijing in 2018 might be a unique example, consider Denver or Baltimore if you're in the U.S., or consider international hotspots like Dubai.
Explore less “sexy” industries: Don’t overlook sectors like oil & gas or insurance. They may not offer free meals, but they can provide a solid foundation in data skills.

5. Second pivot: backwards in salary but forwards in data science

During my first year in London, I did two things:

Looked for data analyst/data scientist jobs as soon as my feet touched the moist soil of Heathrow airport.
Proactively shaped my projects with my tech skills

At work, I took every chance to practice my tech skills.

My team was helping a major bank shift trade data from London to Europe (thanks, Brexit!).

Faced with endless Excel data processing, I jumped at the opportunity to streamline things. I built a Python pipeline that slashed a 3-hour daily grind into minutes and even taught myself VBA to tackle tasks where Python was off-limits.

Outside of work, I was relentless in my job search, aiming high with British tech unicorns like Monzo and Deliveroo. I made it to the middle rounds but never clinched the final offer.

I had my biggest heartbreak in 2019.

I reached the last stage for a data strategist role at Facebook, only to fall short.

I vividly remember that rejection call, huddled in a tiny cubicle, struggling to keep my composure.

“I’m afraid it’s bad news”.

The moment the call ended, tears flowed, but I quickly pulled myself together and walked out as if nothing had happened.

Doubt creeped in: “Maybe I’ll never become a data scientist.” I thought many, many times.

But I kept my head down and pushed through, honing my skills and keeping applying for jobs.

Then, in August 2019, luck finally found me.

I landed a data science consultant role at a SaaS startup. The job was a blend of consulting and tech, which was a perfect match for my skills. Yet I wasn't exactly over the moon. The job was a step down in pay, partly due to my poor negotiation skills.

But I saw this as a crucial step towards my dream job, and most importantly, it gave me the coveted “data science” title.

Take-aways

Leverage your current role: Build the skills you want. Show your team the value of your new skills and gain their support.
Keep applying: Job hunting is often a numbers game. Sometimes, a lower salary is worth it if it means stepping closer to your dream career.

6. Opportunity knocks: the path to Facebook (Meta)

March 2020, my heart raced when I saw the subject line: “Hi from Facebook!”.

I was silently screaming, jumping, and doing impromptu gymnastics in my bedroom.

It was an InMail on LinkedIn about a data scientist role at Facebook, and I wasn’t going to let this chance slip by.

After passing the initial recruiter screening, I had two weeks to prepare for my first interview.

Here’s how I tackled it:

I dove into every FAANG data science interview question I could find, scouring resources like Towards Data Science and Medium.
I found two other candidates on public forums who were also interviewing with Facebook. We became mock interview buddies, talking daily and grilling each other. We bonded over our shared goal and are still friends today.
Facebook provided an interview prep guide, which became my bible. I ensured I could answer every question and explored all the resources they suggested. You can find many resources in my 7 essential skills in DS article.

After six weeks, I finished my final loop with four 1-hour interviews in one day.

Now, the excruciating wait began.

After what seened like a long week, I received an email saying the recruiter would call the next day to "discuss results".

I tossed and turned that night, replaying every detail of the interviews, scrutinizing every answer.

The next day, my phone rang.

“The team is excited to have you onboard!” the recruiter said.

I kept my cool on the call, but inside, my heart was pounding out of my chest. It was the happiest day of my life, especially after countless rejections and the all-too-familiar “We don’t sponsor visas” response.

I felt truly seen and validated.

person throwing confetti — Photo by Ambreen Hasan on Unsplash

Take-aways

Give yourself ample time to prepare: I knew Facebook wasn’t hiring for specific positions but for a pool of data scientists, so the position wouldn't be gone anytime soon. I scheduled each interview round at least two weeks apart
Practice makes perfect: Practice alone and with others. Mock interviews are essential, especially for a FAANG company. My interview buddies, best friend, housemate, and even my ex-boyfriend helped me prepare. I had at least six mock interviews on product sense and three on coding. Let your mock interviewers grill you so the real ones won't.

How I’d learn data science today

If I were starting my journey into data science today, I’d definitely aim for a Master’s in Data Science. With the field now fully mature, countless schools offer robust programs tailored to equip you with the skills you need. While I eventually made it, a structured program could have shaved off four years from my journey.
I’d also dive deeper into An Introduction to Statistical Learning, especially the sections on unsupervised learning and deep learning. These areas are the future of data science, and companies increasingly seek full-stack data scientists who excel across the board.

My top tips for landing a data position in 2024

First off, if you can get a B.S. in Statistics or an M.S. in Data Science, go for it. .
But don’t sweat it if you can’t. Degrees are just one way to get started—they don’t define your potential. What companies care about early in your career are your passion, your eagerness to learn, and your ability to solve their problems. Show them how you can contribute and make an impact.
Find mentors who are already working in the field. They have the latest insights on industry trends and challenges. A good mentor can provide invaluable guidance and fast-track your learning.
Practice, practice, practice. Consistency is key. I dedicated two months to daily practice, from nailing my one-minute introduction to mock interviews with friends and explaining coding problems out loud. There are no shortcuts here—put in the work.
Not sure if data science is your thing? Test the waters in your current job. See if you can apply data science concepts to your projects. You’ll gain hands-on experience and clarity on whether this path suits you.
Remember, you already bring a lot to the table. For instance, my consulting background sharpened my communication and organizational skills. I recently spoke with a civil engineer considering a switch to data science. When he worried about his lack of credentials, I asked, “Why did you choose civil engineering?” He said, “I love building things and solving logical problems.” That’s exactly what data science requires—those are his assets.