Getting into data science without a graduate degree
How I landed one of the most coveted jobs with just a bachelor’s
Back in 2017, I had taken a semester off of college to do an engineering co-op at ExxonMobil’s Beaumont Refinery. I vividly remember wearing my flame-retardant coveralls, steel-toed boots, and safety helmet as I walked along large pipelines and equipment in the 95 degree humid Texas weather. As a midwestern boy with a recently discovered passion for coding, I felt out of place.
If you look at any data scientist job posting today, most if not all of them are looking for candidates with a MS or PhD in data science. Not to mention the laundry list of technical skill requirements.
This is the story of how I traded my hard hat for the sexiest job of the 21st century and 22nd century without even a masters degree.
Start with the basics
Let’s hit rewind.
It’s 2016 and I was starting my junior year as a chemical engineering and economics double major at the University of Nebraska-Lincoln. After getting about a year in research experience at various universities and NASA, I decided I wanted to explore industry.
But chemical engineering was not my only interest. I discovered I loved coding too so I decided to add a computer science minor and enroll in Udacity’s online Data Analyst Nanodegree at the end of 2016. Yup, I loved learning.
Udacity’s program was great. They partnered with top tech companies like Google and Facebook to build online courses to help close the technology skills gap. The Nanodegree has a slightly different curriculum now — it used to include an introductory machine learning and exploratory data analysis courses when I took it.
In 2017, my first term with Exxon consisted of work unrelated to data science. As a co-op, I progressed capital projects and helped support refinery operations. Getting to go out in the field, climbing ladders, and talking with operators was a blast!
I got hands-on experience of what it was like to be a chemical engineer and learned a lot about refinery life. Thankfully, they took a chance on me and I got a return offer to come back to Beaumont for a summer 2018 internship.
While working at Exxon, I continued completing the Nanodegree, finishing it within a year. Every day, I spent 8am — 5pm working and then 6pm — 10pm studying. Then, after my co-op, it was the same schedule but I switched out working for regular university coursework.
Was it rough? Yes. Would I do it again? Absolutely.
Apply learnings
Since completing the Nanodegree, I became more interested in data science than I did about process engineering. Armed with the new credential, I reached back out to Dave, my university’s ExxonMobil recruiting lead (who also happened to be my department head in my first term).
I asked him if there was a possibility to switch my internship to their headquarter office where most of the data science work occurred. I wanted this for four reasons:
- To get real-world experience in data science to see if I would actually enjoy the work
- To learn more about the company and make a global impact instead of site-specific improvements at the refinery
- To set myself up for more options upon graduation — graduates who interned at a site were almost destined to start their early careers at that site
- And to work at the fancy new modern campus — it’s like a futuristic city!
Unfortunately, no switch was available.
So in 2018, I started back in the refinery again as an intern. But this time it was different — I had some data science knowledge under my belt.
From day one, I let Rachel, my new supervisor, know I wanted to get into data science. Though there weren’t any data science positions within the refinery, I was able to convince her to swap one of the three projects I would work on in the summer with a data science related project.
Even better, for that project, Rachel connected me with George, a senior engineer at the refinery that had prior data science experience. It was the perfect pairing. Our high positive energies matched and he eventually became a mentor.
My project? Use machine learning to build an inferential model and improve process control of a multi-million dollar refinery asset. I was so excited that I got to apply topics I learned from the Nanodegree such as supervised learning and hyperparameter tuning.
Halfway through my internship, I had a discussion with my supervisor and my buddy (Exxon pairs us up with a full-time employee for mentorship) on where I potentially wanted to start if offered a full-time role. I said the headquarters office as a data scientist.
I was told the most realistic pathway to get into that role would take at least 5 years. Do a couple rotations at the site as a process engineer, take on a coordinator role, and then maybe, just maybe, be able to transition into a data scientist.
But I didn’t want to wait. I wanted to make an impact on a global scale as soon as possible.
So I needed to demonstrate that I would be more valuable to Exxon as a data scientist than as a process engineer. My strategy? Do a good job and make sure the right people know about it.
Build a great network
I reengaged Dave, who now became the manager of George’s supervisor (how convenient, right?). The project I was tasked with turned out to be one of the big items on Dave’s roadmap. So I kept him in the loop and asked him to help me get exposure to the data science community.
I also took networking into my own hands. After successfully delivering the data science project, I took a chance and scheduled a call with a data science manager at the headquarters. Part of me had a fear of rejection because what busy manager had time for an intern hundreds of miles away at a refinery?
To my surprise, she accepted. She was kind and although she reiterated that my chances of starting as a data scientist were better if I had years of site experience or a graduate degree, she also gave me great insight into how data science worked at ExxonMobil.
I learned Exxon had not just one but multiple data science teams. Some were more established while others were just a year or two old!
A few weeks passed and Dave got back to me. He had reached out to his network and mentioned a data science supervisor was willing to host me for a day at the headquarters. I then asked Rachel if I could go. She was supportive but wanted to make sure I had a plan.
And a plan I gave her.
Upstream. Downstream. IT. I set up a time to talk with as many people that touched the data science space at ExxonMobil. Namedropping the data science manager I previously had a call with helped as well.
On my last week, I drove an hour to a college friend’s apartment in Houston the day before and slept on the floor with a backpack and blanket that night (he was just starting at ExxonMobil and didn’t have any furniture yet).
This was my schedule for the next day:
- 8:00am — 8:30am | Upstream Research Company — Data Science Supervisor
- 8:30am — 9:00am | ExxonMobil IT — Data Scientist
- 9:00am — 10:00am | Value Chain Optimization — Data Scientists
- 10:00am — 11:00am | Real Time Optimization — Optimization Engineer
- 11:00am — 12:00pm | Upstream Research Company — Data Scientist
- 12:00pm — 1:00pm | Lunch w/ Data Scientists
- 1:00pm — 1:30pm | ExxonMobil IT — Data Science Advisor
- 1:30pm — 2:15pm | Fuels & Lubricants — Data Science Supervisor
- 3:00pm — 3:30pm | ExxonMobil IT — Data Science Supervisor
- 3:30pm — 4:00pm | Value Chain Optimization — Department Head
- 4:00pm — 5:00pm | Python Interest Networking Event
Yup, I basically set myself up with a full day of informational interviews and I learned a ton.
Some of the folks I met with asked me about the data science project I worked on, while others shared their work and advice. From this, I realized how large ExxonMobil was and the vast array of interesting projects that were being tackled.
However, one common trend I noticed was that everyone had a graduate degree.
The End Result
On my last day of my second internship, Rachel and my buddy sat me down for an exit interview. There, I would learn whether or not I received a return offer for a full-time position.
And to my surprise, I was offered to a data scientist role at the headquarters!
It felt so great that all the hard work I had put in finally paid off. I was and still am so grateful for all the supportive people that helped align the doors to make this possible.
I later learned that the department head I had talked to on my makeshift super day had called my supervisor to make the switch. One of the reasons was because I brought something new to the data science team that they didn’t have — my whole 8 months of intern refinery experience.
Since starting at Exxon full-time in May 2019, I have gotten the opportunity to work on a variety of projects from trading recommendations to pricing guidance and equipment optimization, helping to capture $15M+ each year.
Fun fact: In one of my new projects, I helped deliver a global product that expanded on the work I did as an intern. Guess who was one of the first product owners? — It was George. It truly is a small world.
Takeaways
Landing a data science position boils down to ensuring you understand the fundamentals, applying that knowledge in your current role, and connecting with the right people that can support you and open the right doors.
All of these start with you. You are the driver and your determination is the fuel that takes you to your destination.
Remember, your background is unique. Use that to your advantage and surround yourself with those that recognize that. And when the right door opens up, make sure you’re ready to walk through!
Lastly, be grateful and kind to everyone you meet. You never know when you’ll need them in the future.
Are you trying to get into data science? Message me on LinkedIn! I’d love to be a part of your journey.
