We created CommunitiCash because we believe that there must be a better way of insuring people's livelihoods. The United States has a large amount of people living paycheck to paycheck, for whom even small disturbances in their finances can lead to large consequences. In the status quo, there exists little recourse for such individuals, which is where CommunitiCash comes in.

What it does

CommunitiCash uses machine learning to analyze the financial histories of its users to aggregate them into Groups with the lowest risk for all members involved. Our algorithm predicts the future of a user's financial situation, and uses this information to connect them with others who can ensure no one falls behind. With Groups being limited to 20 members each, members are encouraged to do their part in supporting one another.

How We built it

CommunitiCash's machine learning algorithm was developed using Python, particularly with the NumPy mathematical library. After it was run on the original set of data that was provided by Vitech, the new data was hosted on Firebase, where it could be accessed by our front-end application written in JavaScript.

Challenges We ran into

One of the biggest challenges with this project was ensuring that the algorithm was accurate, especially when dealing with outliers (very high/low income users). Another challenge was transferring all of the data to Firebase, as the original Vitech data was a CSV, and Firebase's database stores JSON objects. The CSV to JSON converter we had planned on using failed on us during the night, so we were forced to convert the data ourselves.

Accomplishments that We're proud of

Knowing that our application is predicting the future to potentially help people is something we're very proud of. Not to mention, finishing the application at all despite the problems we encountered (particularly with the CSV to JSON conversion) is pretty nice too.

What We learned

We learned how to apply machine learning and probability to financial data sets, something that we most definitely could use again in the future. (We also learned that just because a CSV to JSON converter worked two weeks ago, doesn't mean it will work when you need it to...)

What's next for CommunitiCash

Our plan for CommunitiCash is to aggregate enough recorded and predicted financial data on users to be able to present to banks and money lenders, to prove that Groups are stable enough to be able to take out bigger loans and still have the capacity to pay them off. This benefits everyone involved as Group members are able to have more money for carrying out their desires, and money lenders can be more confident in knowing that they will get a return on their investment.

Built With

Share this project: