PennyPro

INSPIRATION

Ramp, Duolingo

PROJECT OVERVIEW

Check out our GitHub repository’s README to learn more how this project works!

HOW WE BUILT PennyPro

Plaid API - Allows us to interface with users' banks to provide visualizations of the user's current financial breakdown.
SvelteKit - Front-end framework
DaisyUI (TailwindCSS) - CSS component library
Go + Pocketbase - Localhost RDBMS
Python + ML libraries - User-data processing and text summarization model to train a financial advice LLM
FireBase - User authentication/login management

CHALLENGES

Time was the biggest challenge for us. Unfamiliarity with various technologies required our time to be dedicated toward research rather than implementation. Difficulty integrating certain APIs lead to more time restraints. Interfacing different elements of the project also proved to be a struggle. Generally, a lack of time forced us to abandon planned features, as the wireframe for the project required more organization than anticipated.

TEAM ACCOMPLISHMENTS

Thanks to our extensive pre-planning, we were able to focus on our project’s MVP without any issues. We were able to construct the foundation of the integration tools needed for the project. Along the way, we were learning new languages, frameworks, and tools many of us were not familiar with.

WHAT WE LEARNED

Relational state transferred APIs and API design practices
Svelte-Kit framework
Gained a professional understanding of utilizing an Entity Relationship Diagram

WHAT’S NEXT FOR PennyPro

Connecting the backend and frontend.
UI/UX optimization on frontend
Backend logic optimization
Train a conversational LLM to provide an interactive learning option. (Currently running a summary/tokenizer model to grow a dataset of summarized financial literature)
Incorporate live market values on various items to improve the user’s budgeting, providing discounted items (Ex. eggs are currently $2 cheaper here)