*As the summer approaches and many students are preparing for internships at FinTech companies, we thought it would be appropriate to ask Shantanu Gangal (WG’15) to explain his internship at ZestFinance and what advice he has for future interns.*
ZestFinance is a FinTech startup founded in 2009 and based in Hollywood, CA. They describe themselves as “Google-style machine learning meets CapitalOne-style credit scoring.” This post captures my experiences and learnings over the summer at Zest.
What Zest does
Zest applies big data analytics to credit underwriting. Through their proprietary machine learning algorithms, Zest analyses “thousands of potential credit variables – everything from financial information to technology usage – to better assess factors like the potential for fraud and the risk of default.” The entire application is online and the credit decision takes only a few seconds. If approved, borrowers could get funds the next day. Zest provides credit to “roughly one in four Americans who don’t have access to traditional credit cards [at rates lower than payday lenders”](http://blogs.wsj.com/venturecapital/2012/01/19/the-daily-start-up-zestcash-nabs-73m-to-disrupt-payday-loans/) but much higher than credit card rates.
Zest has grown fast over the last six years, recently raising their third round of VC funding from Peter Thiel [bringing total funding to ~$112m] (https://www.crunchbase.com/organization/zestfinance). Prior investors in Zest include Lightspeed and Matrix, among others.
My role during the internship
Zest’s proprietary model relies on data provided by applicants and other external databases. Using this data, the model spots applicants that are lacking either intent or ability to repay a loan. However, we wanted to go beyond reported data. We decided to investigate if we could identify behavioral factors that would add color to an applicant’s profile. I classified applicants into two categories:
**Readers:** the folks who were very diligent about what they were borrowing
**Speeders:** people more cavalier about the process as a whole
By testing the actual performance of these applicants over historical data, I was able to establish that Readers were a much better segment to lend to. We further established that the result was statistically significant even in the presence of existing information about the borrowers. It is a moment of pure joy when data validates a gut hunch. Exhibit A illustrates the benefit of the new classification (scrubbed data).
I extensively used SQL (database querying), R (statistical modeling) and Google Analytics (tracking) over the summer. In addition, to make a mark among the wickedly smart PhDs at Zest, I needed to lean heavily on the Wharton STAT core course (who would have thought?!). I would strongly recommend being on top of these materials if you are interested in working on a similar project.
A window into startup culture
This was my first experience working at a startup (after jobs at BCG and Blackstone – both relatively large and global companies). I noticed that everyone got their hands dirty with coding, testing or analysis as the case may be. While there was a reporting structure, each one of us was empowered to make all decisions regarding our work streams. Formal check-ins with managers were limited to only about an hour each week.
In my role I needed to apply a wide range of analytical and statistical tools and there was a lot learnt by asking the team around me. Being able to walk up to people with a question multiple times within a single hour was what made it a very efficient and collaborative learning experience.
Zest was also my first experience with fantastic startup perks – wearing jeans to work and playing ping pong during the lunch break!
My experience at ZestFinance was extremely rewarding. Coming from a finance background, it was an excellent segue into the startup world. I learned about startup culture, had a chance to apply the statistics and data tools from MBA classes and most importantly, met a bunch of smart folks working on interesting problems in lending over the internet. Good luck to all the students interning at FinTech companies this summer!
Shantanu is a second year MBA student at Wharton. Earlier, he graduated top of his class at IIT Bombay (CSE). Shantanu is very interested in tech-enabled credit. This past summer, he worked at ZestFinance. In the following semester, he completed an Independent Study on Lending Club’s portfolio and is currently working on one with Prosper. Prior to Wharton, he worked at Blackstone & BCG.