COMPOUND

Sign in
Sign up
Landing
Leaderboard
Profile
Resources
Camera
Home

Inspiration

Apps like Venmo and BeReal offering a social element to daily life and continuing to be way of our spending habits in college.

What it does

Compound reads receipts with the camera, allowing users to track their spending easily and compete with friends to save the most money. It brings the community together to fight overconsumption and fosters a playful and healthy environment!

How we built it

We built the project using a combination of Perplexity's API and Pytesseract for optical character recognition (OCR) to extract text from receipt images. Once we obtained the text, we used Perplexity to categorize and summarize the receipt data into predefined categories like Food, Transportation, and Shopping. To visualize the results, we integrated Matplotlib to create pie charts displaying spending breakdowns. The process involved fine-tuning API calls and ensuring smooth data flow from OCR extraction to categorized visualization.

Challenges we ran into

One of the major challenges we faced was handling large blocks of text extracted from the receipts. Perplexity sometimes returned incomplete or malformed JSON responses when dealing with larger inputs, which required us to implement extra validation steps and error handling. Additionally, cleaning the extracted text from pytesseract to ensure accurate categorization posed some difficulties due to inconsistencies in the OCR output. Despite these challenges, we were able to iterate and improve the workflow for better accuracy.