Three Years to FinTech

March 2021

My job for most of college — and the profession where I began my career — was public accounting. I did a handful of internships in corporate finance and regulatory accounting, but public accounting ended up being a great opportunity to really put the academic theory to practice. Accounting is quite systematic, but as no two accounting systems are the same, it hasn't yet been completely automated. The bulk of my job was serving as an intermediary between these various systems and the other members of our audit team.

Download a report.
Check if there are any new updates.
Document your findings

It felt like the kind of work that a computer could do — work that a computer should do. I wasn't sure if it was possible, but if it were, I knew I'd save a ton of time.

During the Spring of 2017, I got the opporunity to subcontract on a federal audit with a larger firm whose office was across the street from us. The large auditors were required to bring on a certain number of local accountants, and our firm had partnered with them for over a decade to send a few of our auditors over to help with the year-end audits of whatever federal clients they had. I applied, and I was able to join the federal audit of their federal client: the US Department of Commerce and Patent and Trademark Office. The year end schedule was particularly grueling, as our contract explicitly required a minimum amount of butts-in-seat-hours each week, regardless of how much work was available. I spent my downtime learning how to program.

The first task I worked on involved web scraping, email, and working with spreadsheets and PDFs. It was procedural, straightforward, and boring, so naturally I felt drawn to Automate the Boring Stuff with Python. I opened the table of contents and saw several chapters which were directly related to my task, and since the HTML version of the book was available for free, it was enough for me to get started.

The task took me about four or five minutes to do manually, so the first time my program ran, I assumed it would take a similar amount of time. After running my script for the first time, I went to grab coffee, but before I reached the break room my phone got a notification that the case files were updated with my report. I reviewed it to make sure everything was correct, and then signed off the audit procedure. It took less than 15 seconds.

While walking back to grab coffee, my mind was thinking about how significant this was. Saving a few minutes on each client wasn't a significant amount, but a 90% reduction in time, extrapolated over every audit procedure would be massive. I had significantly underestimated how much of an impact software could have on financial services. In the future, everything would be fully automated, interconnected, and available on demand, and I decided that day that I wanted to work on making that future a reality.

Before I found Automate the Boring Stuff, I discovered lots of resources for people who wanted to become self-taught developers. I thought I could learn how to write more complicated scripts on my own, but I was skeptical about my ability to learn independently. I had studied Spanish and Japanese throughout high school and college but never became fluent. I saw programming languages in a similar vein, and assumed that I would need to immerse myself entirely in a software rather than learn at a leisurely pace. Even if I could learn on my own, I felt that getting proper instruction would expedite my learning process, and since I had just come from graduate school, I started looking at graduate programs in computer science.

There were several universities in the DMV area which I was considering. Most of the programs were structured as 60 credits over 2 years, or 15 credits per semester. But because I lacked a technical academic background, there was an extra semester's worth of prerequeistes I would need to complete. I had about 18 months of savings and I was willing to quit work and devote myself entirely to my studies. I wanted to take 20-25 credits a semester and complete this in about 12 months, which would give me a few months to find a job. I spoke with the admissions directors of these programs and asked if I could load up on credits to complete the program quicker, but none of these programs allowed me to take such a large classload, so a graduate degree would take at least two years. Add up the forgone wages and tuition, and this became a six-figure financial decision. Before I committed, I wanted to explore other options.

My roommate at the time happened to be making a career switch of his own, and was in the middle of a programming bootcamp that summer. I asked if I could sit in on one of his classes, and saw that it was 8 hours every day, which was about how much time I was hoping to spend if I were taking a 25 credit workload in school. His program was about 15 weeks long and exclusively taught JavaScript, a language he said was good for developing web applications, which was what he was interested in doing.

I started to consider what programming languages I needed to learn. I had started with Python and enjoyed the ease with which I could automate manual tasks, but I liked the idea of working with web applications so that I could connect different systems and make them accessible to anyone from any device. I looked through the career pages of companies I was interested in working with and made a listing of the technologies they listed in their job postings.

It seemed like 4 languages were prevelant across the companies: JavaScript for web apps, Java for mobile development and banking systems, C++ for high performance trading, and Python for data science and analysis. I decided that if I could find a program to teach me 2 or 3 of these, I would be in a good position to get a job after graduating. My roommate's program taught one language over 15 weeks, so it seemed like an ideal program for me would be closer to 30 weeks.

I happened to read this post on Hacker News about a programming bootcamp which seemed to meet my criteria: 8 hours a day, 26 weeks long, and they taught JavaScript, Python, and C++. Most importantly, they had an Income Share Agreement which, while more expensive than debt, gave me a risk-free guarentee that I’d be learning the skills needed to actually develop professional software — something that wasn’t guarantee with a traditional university degree. The worst case scenario would be 6 months of foregone wages to learn something new, and then I would go back to public accounting and not owe a penny. And unlike most student loans, I only had to repay the ISA while I was employed. The premium was steep, but the flexibility more than made up for it.

I completed the precourse application in November, which was also a positive sign for me. It was self-paced and designed so anyone could learn how to complete it with adequate effort, filtering out students who weren’t necessarily committed to the program. I was accepted the week of Thanksgiving to start in the December batch, but after tellling my family over Thanksgiving dinner, I pushed it back to January because of how adamantly against it they were. I got a similarly visceral reaction from talking to my grandparents when we visted for the holidays in December, and ultimately pushed my start date back further to March.

I always went to my parents and grandparents for advice — they are sort of my personal board of directors. 99.99% of the time, I’m missing something obvious and their advice is right on the money. They're normally able to see the inconsistencies and biases in my thinking, but this was the first time I really felt like they were the ones missing something obvious.

There’s less money in IT.

You’ve already got an accounting degree, why do you want to get into technology?

It’s risky.

You should get an MBA.

They didn’t seem to grasp the financial part of financial technology, and the fact that they kept couldn’t understand the difference between IT and software development was somewhat concerning. The MBA was interesting, but what value would I get out of that so early in my career? Unconvinced, I put in my notice at the end of January, left on Friday, March 2nd, and started my program on Monday, March 5th.

I finished the bootcamp in the middle of September and landed my first developer job in October. I wasn’t accepted to any financial technology companies, but I joined a small consultancy that was great for helping me ease into professional software development. I primarily worked on the frontend, but I got to dabble in Python and had the opportunity to learn cloud infrastructure as well. Overall, it was a great learning opportunity — and my starting salary was 30% more than what I was making in public accounting.

About a year later into the role, I was able to get onto a larger client of ours. They were a billion-dollar commercial real estate company, and contracted with us to do some really cool visualization and 3D rendering work. It was awesome — except a few months later when COVID-19 hit and commercial real estate took a nose dive.

Around summer 2020, I felt like I was ready to move into a financial technology role. I applied for a job at Capital One, went through the interview, and passed!

Well, kind of.

After my last round of interviewing, I got a call from the recruiter saying they wanted to hire me, but they just filled the job, and they couldn’t hire me at that time. They said they would keep me in their system for six months, so if anything opened up, they would call me and I wouldn’t have to re-interview. I thought that was their way of letting me down easy.

Unfortunately, this coincided with my client issuing a stop work order on our project, and a few weeks later, I got let go. Ironically, that was my last month of ISA payments. I called the Capital One recruiter who reaffirmed that there still weren’t any opportunities, but that more may open in the new year. I wasn't willing to count on it, so I started searching for new roles.

I was in a uniquely fortunate situation. We had just refinanced our condo, and both my ISA and my wife’s student loans were paid off a few months before this happened. We would be able to live off her income without dipping into savings, and she was supportive of me being more selective with my job search so I could focus on interviewing with financial technology companies. Because COVID pushed everyone remote, I was able to apply to a broader range of companies which would have been geographically unavailable during normal times, so I wanted to take advantage of it.

At the same time, I knew I couldn’t wait forever for a dream job. I made a six-month plan in which I’d be extremely selective in the first two months, then expand my search to anything in the technology or financial space, and after four months, I would apply to any job I was qualified for.

After the first few weeks of job searching, I realized that I was going through job boards faster than jobs were being posted, so I couldn’t spend more than few hours applying for jobs each day. I got back into running, but also decided to finish my CPA licensure, as I had met all the requirements other than passing some of the exams. My schedule was to wake up, apply to jobs, run, study, eat, then sleep.

Over the next 6 weeks, I passed the remaining CPA exams and was officially licensed in early December, which was just in time for me to start expanding my applications to financial and accounting related jobs. In the middle of December, I applied to a very interesting apprenticeship program at Microsoft, which I thought would be a great opportunity.

A few weeks later, just before Christmas, I received and accepted an offer from the U.S. Tax Court as a support engineer, troubleshooting tickets from users and helping to develop new features for their case management system. I was making less than I was making as an accountant, but it was a job that I would be able to learn a lot from, so I was fortunate to get a job before the new year. I was slated to start the first week of January and I withdrew myself from consideration from all the other jobs I had applied to — except for 2.

The first exception was the Microsoft role I applied to back in December. After I applied, there was no point of contact to reach out to, so I didn’t know how to withdraw my application. In the new year, they invited me to interview at the end of January, and I received an offer a few days later to start in March. It was a 16-week Technical Program Management rotation, and I would be given the opportunity to apply to a full time role afterwards. The program was remote during the pandemic, but I would have to move out to Seattle once the offices opened up.

While I was looking at apartments in Seattle, I got a call from the second exception in February — Capital One. I met with the engineering manager and they followed up a few days later with an offer. I was surprised that they reached out, but I was just within the six-month window from when I had originally applied. Had they waited until March to open the job, I would have had to go through the application process all over again. But unlike the Microsoft role, this was a full-time role and their office was 4 miles away from me. And most importantly — it was FinTech.

At Capital One, I get to work on fascinating financial technology at a scale I could never have imagined three years ago. FinTech is actually an amazing opportunity to do good in the world. Every individual is involved in a global financial system, and technology provides a way to make this system more transparent, accessible, and affordable. Software enables us to interact and transact with each other whenever we want, however we want. Over time, software will whittle down the barriers between people, and expand the wealth of everyone, not just a select few.

Some of these changes occur suddenly and completely revolutionize the way financial services are performed, whereas other changes will be slow and incremental. But regardless of how FinTech changes the world, I know it will be in a meaningful, tangible way. It's not open heart surgery, but it's also not optimizing ad impressions. It lies somewhere in between, and if I can spend my professional career working toward moving the needle forward, I would be content to retire knowing I played a part in something good. I contributed to something bigger than myself. I made a difference.