Bigfin

BB&T Data
Capital One Data
About Page

What it does

Bigfin is a private financial watchdog application, allowing consumers to make informed decisions about their banking. Using complaint data from the Consumer Financial Protection Bureau ("CFPB"), which was established to regulate banks following the 2008 financial crisis, the app collects quantitative, qualitative, and geographic data regarding large and medium-sized banks. Bigfin analyzes the data using natural language processing and basic statistical models to determine the risk and reliability of a given bank. As a bonus feature, users have the ability save banks of their choice, and receive daily notification alerts when there is a change in risk.

How we built it

Anubis built the front end, including the design of our UI, all of the charts, and severity of the complaints. In the process, he used javascript, data visualization libraries, and a node server. Neil completed the back end, scraping data from the CFPB, using Natural Language Processing to analyze the complaints, and finding bank data points using their public profiles. After hours of late-night work, the two sections were merged successfully around 4 AM.

Challenges we ran into

Using data visualization libraries for the first time (moving from d3js over to chartjs)
User data persistence (user session stays when site is reloaded)
Severity classification with NLP Library (switched from attempting Google Cloud to nltk)

Accomplishments that we're proud of

During the development of Bigfin, we successfully used a data visualization library for the first time! In addition, this was Neil's first ever hackathon and we are proud to have built a useful product together.

What we learned

We learned how to create a data visualization model of AI-analyzed data! Some specific libraries that we learned to use for the first time were nltk for natural language processing, chartjs for creating interactive graphs, node servers, and SAS for advanced CSS.