Inspiration
We noticed that while banking apps provide transaction histories and charts, they rarely teach financial literacy or adapt to individual users’ habits. We wanted to build a tool that goes beyond static dashboards, one that acts like a financial mentor, helping users understand spending psychology, identify inefficiencies, and build healthier money habits.
What it does
Savvify allows the user to either login/signup and based on their details it pulls their credit card details and displays their credit scores along with a breakdown of their expenditure. It also contains a resource breaking down modern financial terms and how to build credit.
How we built it
We built the frontend using React + TailwindCSS for a smooth, responsive design. The backend runs on Flask (Python), connecting top the anonymized financial data and generate insights.We used: OpenAI API Call Framer Motion for smooth dashboard and card animations. All data updates dynamically using local caching for performance.
Challenges we ran into
At first we tried to run a model from hugging face in order to build our chatbot, however, we failed since the graphics card that we were using was not compatible with the specific version of PyTorch that we had used. So we had to switch to using an OpenAI API call and since we had to upload our codebase on GitHub but apparently OpenAI considers it as an 'API leak' so the chatbot is unable to function as of now.
Accomplishments that we're proud of
We're proud that we were able to build something that could be applied to a real situation and benefit the community and customers overall.
What we learned
We liked that we went head first into problems without second thoughts and that we were able to pivot every time we hit a roadblock and didn't spend too much time reminiscing about what went wrong. Moreover, all the debugging that we had to do was more intense than any of our coursework and the very fact that it had a huge learning curve was also what excited us.
What's next for Savvify
In the future, we plan on using LLMs locally to understand how customers of different ages would understand certain financial concepts (eg: If the customer were a young adult they'd likely be more engaged by a different way of explaining as opposed to an elderly). Moreover, we hope to have gained more experience working with GitHub which most of our team hadn't before so hopefully we'll work on that in the future. Also, perhaps we could work on making our UI more 'aesthetic' per say. Either way, we're confident and glad to have built Savvify in these 24 hours and we plan on bettering it as we become more learned programmers.
Log in or sign up for Devpost to join the conversation.