Inspiration
Splitting a restaurant bill can be frustrating when people do not remember who ordered what. SplitPay was created to make group payments faster, easier, and more accurate.
What it does
SplitPay lets users scan or upload a receipt, review the extracted items, assign items to people, calculate each person’s share, and send payment requests through PayPal Sandbox.
How we built it
We built SplitPay with Next.js, TypeScript, and Tailwind CSS. Gemini parses receipt images, while the app calculates totals in code to keep the split accurate. PayPal’s Invoicing API is used for sandbox invoice testing.
Challenges we ran into
We had to handle camera permissions, unreadable receipt images, Gemini model availability, exact cent-based calculations, and PayPal Sandbox configuration.
Accomplishments that we're proud of
We are proud that SplitPay can turn a receipt image into editable items and calculate accurate individual shares. We also successfully connected the app to PayPal Sandbox invoices.
What we learned
We learned how to integrate AI image processing, build a multi-step user flow, work with payment APIs, and design reliable money calculations.
What's next for SplitPay
Next, we plan to improve receipt recognition, add AI-powered item assignment, support PayPal webhooks, and improve the overall mobile experience.
Built With
- gemini-api
- next.js
- paypal-invoicing-api
- paypal-sandbox
- react
- tailwind-css
- typescript
