What MIT’s FinTech Hackathon Says About Big Banks

3 Reflections on a weekend of hacking

MIT hosted a FinTech Hackathon this weekend. About 140 people gave up their weekends and in 36 hours had to come up with a new product. The overall winner received $5,000. While people were free to work on whatever project they wanted, there were four sponsored hacks that had additional prizes.

I signed up to meet some people to get a better feel from participants and sponsors what they view as the biggest opportunities and needs are in FinTech.

Here are 3 reflections on the weekend:

  • Thoughts on the winning team
  • What was in it for the big banks?
  • Lessons from the finalists

The Winner

Switch, a team from Cornell Tech, won the grand prize. They built a prototype for an app that would offer freelancers personalized loans to compensate for income shortages.

The Winning Team

The problem they presented is very clear. The US (and world) economy is increasingly moving to freelancing, but freelancing comes with tremendous income instability and no benefits. Switch addresses this by offering loans to smooth income shortfalls, and in the future will add insurance quotes to its offering to help freelancers get access to insurance.

To verify a user’s income, their plan is to integrate with freelancer marketplaces (e.g. Upwork, Uber, Lyft). From the presentation, it sounds like income would be the main data collected to evaluate risk. Switch would then integrate with Even Financial’s API to get personalized quotes from online lenders based on the risk profile of that particular user.

Switch were worth winners. They had a great demo, a well defined problem, and an interesting solution. It also sounds like they plan to build out and release their product.

Thanks to the work I’ve done for TycoonApp, I’ve thought a lot about the freelancer market. A few initial questions for Switch’s strategy:

Customer Research

  • The presentation started from the assumption that these problems exist. Freelancers must be doing something in response to these challenges. How do they currently finance shortages? How do they currently insure?
  • Similarly, has Switch talked with full-time freelancers to understand their perspective? Maybe instead of a loan when there’s a shortage, they’d be willing to pay extra to find more work? Or maybe invoicing, collections, and taxes are the main challenge.

Target Market
The overall freelance market is big, but I doubt the number of people who earn all their income through marketplaces is particularly big.

  • A cursory search shows <200,000 drivers on Uber, and ~100,000 on Lyft. Most Lyft drivers are also signed up for Uber, which puts an estimate for Uber and Lyft at <250,000 drivers. Do enough freelancers work through marketplaces to make this a viable option?
  • I still get most of my contracts through referrals, even though I’m on a series of platforms. How many of the freelancers work exclusively in these marketplaces so that Switch can get an accurate view of their income?

Data Sources

  • Are there other ways to get access to the income data? The Switch team had mentioned possibly integrating with bank accounts. What else is there?
  • There is tons of other data Switch could collect on their freelancers to help determine the risk. What other information could they submit to Even’s API?


  • A lot of companies are trying to target freelancers right now. Given that you’ll be sharing profits with the lender, Even Financial and yourself, how much margin is left?
  • How high are customer acquisition costs?

Freelancers are a growing space and face a host of challenges. I’m excited to see where Switch goes moving forward. They were able to build an amazing demo in just the weekend.

What the challenges said about FinTech and Big Banks

  • CitiBank sponsored one challenge on how Citi could understand and use the connection between physical and financial health.
  • TD Bank sponsored two challenges, both around improving the service levels customers received in banking.
  • Even Financial sponsored the final track for the best use of its API around real-time, targeted loan rates.

After speaking with Citi and TD, it became clear they’re both thinking a lot about personalization. How can banks provide a more personalized experience to its customers? What insights can they find about clients that nobody else might have? How can this improve their relationship with clients? How can this increase customer loyalty to a bank? How can this increase the chance of selling additional products?

Citi’s and TD’s challenges were very different, but where they connected was this need to find additional insights relevant to customers. It felt like both had come to MIT to crowd source ideas around this.

TD was interested in two different areas. First, how to improve the way customers interact with the bank, and second innovating the way banks interact with customers (via banks and call centers) to improve the customer experience. The first one explored new technologies (apps, Alexa, algorithms) that could deliver insights directly to clients, and the second one was all about improving the existing experience.

It felt strange to me that these challenges were separated in this way. The direct customer experience through technology and the experience in the bank are inherently connected. Maybe this was only done because we had 36 hours? Or maybe banks are struggling with large-scale change?

Rather than pursue large-scale change, the banks were looking for targeted solutions that move things step-by-step in the right direction.

Lessons from Finalists

There were three finalists from each of the challenges. Here are just some rapid reactions to what I saw:

  • The best teams did their own user research (standing outside starbucks, cold calling potential customers, surveys to fiends, etc.)
  • The best teams also had a strong user persona and presented their results through this lens
  • Personalization was everywhere
  • So were “AI powered insights” and “Algorithms.” Lots of lofty promises were made on what these could accomplish and how easy it might be
  • Personalization and insights were often presented as a response to “the inherent distrust of financial institutions”
  • I loved how many different types of solutions were presented. Chatbots, chrome extension, Alexa app, and mobile apps all made the mix.
  • Design and story was more important than technical accomplishment. The very best teams did all three (design, story and tech), but well-designed mockups and a good story was enough to get many teams to the finals.

I had a great weekend with MIT FinTech. Thanks for hosting!

Show your support

Clapping shows how much you appreciated Daniel Heyman’s story.