Inspiration
Need for Accessibility
- Financial advisors typically seek high-income clients, widening economic disparity between wealth classes
- The growing digitalization of banking services excludes key groups such as seniors
Need for Education
- Rates of financial literacy amongst adults in developing countries are extremely low
- Financial literacy is not prevalent within school curricula, resulting in uneducated youths
What it does
Teaches --> provides a beginner-friendly platform for all levels of financial literacy to engage with and guides users with comprehensive financial explanations
Recommends --> points users to potential courses of action for improvements in a wide variety of goals such as budgeting
Reads Transaction History --> reads transaction history in a spreadsheet format to gain insight into purchasing patterns
How we built it
Backend --> wrote a Flask server with Vertex AI Gemini API that takes in POST requests from the frontend, ultimately returning the AI-generated response
Frontend --> built a react app with a display component and an input component that sends POST requests to the backend
Challenges we ran into
User Privacy
We wanted to strike a balance between the personalization of the AI and user privacy as it raises ethical concerns when a bot possesses unlimited control over transaction history. To mitigate this, we decided on the implementation of an opt-in/opt-out system that would give the user a choice based on their preferences.
Accomplishments that we're proud of
We were extremely proud of being able to produce a finished product despite neither of us having a proper tech background. Additionally, our flexibility throughout our learning journey demonstrated an unmatched level of learning and resilience that we hope to carry onto future hackathons. :)
What we learned
Putting together a formal project with our programming knowledge is relatively new ground for us, making fiNavi a huge leap. We were able to learn a lot about formal project organization as well as navigating Google Developer tools.
What's next for fiNavi
Integration into Banking Applications --> adoption by existing banking services seamlessly with current user interfaces
Deeper Model Training --> train the model with a combination of new and existing datasets to help improve its knowledge base and recommendations
Quality Improvements --> implement speech to text to improve accessibility for senior users and an insight tab to provide overall investment portfolio insights