Finding FishE

Landing Page
Dashboard
Receipt Verification

Inspiration

Inspiration for our project came from a desire to level the playing field between small business owners and larger companies. We realized that small businesses have a lack of resources and tools to manage and monitor their expenses, while larger companies have the resources to do so. To address this issue, we created a web app designed to help small business owners by providing them with the ability to determine if receipts represent suspicious spending on their company cards. Our solution aims to offer small business owners a yet powerful tool to safeguard their finances and promote transparency within their operations. By equipping small business owners with this technology, we hope to help small businesses thrive.

What it does

Our website allows the small business owner to upload a configuration document, which could be a sheet detailing the policy on spending with the business card usage, or a straightforward list outlining the allowed and not allowed expenses. Following this document, business owners can then send in receipts which are sent to our backend. In the backend, by utilizing NLP we are able to determine if the receipts align with what the configuration document set forth by the business owner. This approach streamlines the process saving the business owner time and resources while also saving the money

Challenges we ran into

Some challenges that we ran into were connecting the front and back end together. No one in our group had utilized flask or app routes before so it was especially rough trying to learn and implement those technologies and concepts over the course of two nights. Navigating HTTP requests and responses, along with understanding the correct way to format our data for sending it back and forth between the front and backend proved to be a difficult task.

Accomplishments that we're proud of

Our biggest accomplishment is that we got flask and the app routes to work. Enabling us to send jsons back and forth between the front and back end. Another accomplishment that we are proud of is how well everybody on our team worked together. At first we didn't really communicate our ideas with one another, but as we got to know each other we all started to bounce ideas off of one another or even just to ask for help. Everyone played their role on our team and were all so eager to help one another.

What we learned

We definitely learned some technical skills, whether that be from learning what a json is all the way to how to use and implement Supabase. Although we struggled a lot with connecting the front end to the back end, that is also where we learned the most about app routes, HTTP methods, and jsons. We also learned how to integrate python and Javascript programs together.

What's next for Finding FishE

First we can create unique profiles for each business owner. Once doing so, we can utilize the receipts that they accept or deny into a training set that we can train a ML model on to better predict receipts they are likely to accept/reject. Furthermore an introduction of a more intuitive userface so that the tool is more accessible even to people that are not familiar with technology.