Ready, Set, Mortgage!

An interactive wizard to help new potential homeowners determine if they're ready for home ownership based on their finances, and provide helpful feedback using Generative AI!

Comment

Inspiration

We were very heavily inspired by the Fannie Mae challenge statement, as it's a kind of problem that we have all faced in one way or another. Even though we're not out trying to buy homes as young adults, many aspects of personal finance can be incredibly frustrating. It can feel like wading into a sea of numbers, with no idea where to even go. With this tool, we hope to change that for the process of taking out a mortgage. While not everyone may be shouting "Ready, Set, Mortgage!" like we are, we hope that this tool will take the stress out of this confusing process and help potential future homeowners effectively focus on getting themselves ready for home ownership.

Another big inspiration for us to pursue the Fannie Mae challenge was the enthusiasm and excitement that the representative we spoke to expressed about the possibilities available through this challenge. While mortgages and home financing aren't typically regarded as "exciting" topics, it was incredibly clear that they were very passionate about the work they do and about using data and emerging technologies to improve the home financing experience. His energy really propelled us into taking a crack at this challenge and bringing an air of fun and excitement into the mix.

What it does

Our website is an interactive wizard. Users input a few details about themselves, such as their income and debts, as well as some details about the home they want to buy, such as it's appraised value and the size of their down payment. We then use those to calculate and compare four main criteria to assess loan readiness:

  • Credit score

    • Ratio of loan amount to house value
    • Ratio of total debts to gross income
    • Ratio of estimated mortgage payment to gross income

From there, we present this information to the user. The report tells them whether they're ready for home ownership, and if not, which criteria need improvement. We then employ generative AI to provide personalized financial suggestions to the user based on their individual financial situation. It can sometimes be quite hard to understand how each of these criteria work and how they can be improved, but the AI component truly shines in delivering effective and meaningful advice.

How we built it

The frontend is a React application that uses Bootstrap components. It communicates with the backend using a REST API that we implemented in Python with FastAPI. We also leverage OpenAI's GPT 3.5 Turbo API to provide AI-powered financial recommendations.

Challenges we ran into

One of the biggest, but oddly rewarding, challenges was understanding the mortgage approval process. The whole reasoning behind this project surrounds the fact that, for many people, understanding the factors that go into mortgage approval can seem confusing, and that's a position we were in at the start. It was very important to us that we were actually understanding what the criterion meant, rather than just performing mathematical operations, because this application is focused on the story of a potential new homeowner, who needs more meaningful feedback than just numbers on a screen.

We also ran into a few challenges around connecting the backend and frontend, namely some oddities with cross-origin resource sharing and ensuring both sides were talking to each other and providing each other the correct data.

Accomplishments that we're proud of

We're really proud of the project as a whole! None of us knew each other in the slightest when we first sat down to work on this project, but we were able to work together and create something we can be proud of.

What we learned

We learned a lot during this hackathon! We learned about implementing APIs in FastAPI, and how to effectively define schema to communicate efficiently between frontend and backend. We also learned how to work with the OpenAI API and how useful (but tricky) working with generative AI can be.

What's next for Ready, Set, Mortgage!

There's so much more potential for Ready, Set, Mortgage!. Continuing to refine the UI, potentially with feedback from users, would be a great first step. Improving the generative AI integration could also be an avenue with a lot of potential, allowing users to ask further questions in response to the advice from the AI model.

Built With

Share this project:

Updates