PayPilot AI – AI-powered Payments & Cash Flow Assistant
Inspiration
Small businesses often struggle to manage invoices, track incoming payments, follow up with overdue customers, and maintain healthy cash flow.
We built PayPilot AI to bring invoice management, AI-powered financial insights, and digital payment capabilities together in one platform.
What it does
PayPilot AI helps small businesses manage their payment operations through:
- Financial Dashboard: Displays total invoice value, collected payments, pending balances, and overdue invoices.
- Invoice Management: Tracks customer invoices, due dates, payment statuses, and outstanding balances.
- AI Business Assistant: Uses Google Gemini to analyse invoice records, identify overdue accounts, highlight collection risks, and recommend actions.
- PayPal Integration: Uses PayPal Sandbox APIs to create payment orders, support checkout approval, and capture payments.
- Payment History: Records transaction details, including order IDs, capture IDs, gross amounts, PayPal fees, and net amounts.
- Invoice Reconciliation: Supports linking payments to invoices and updating outstanding balances.
How we built it
We developed PayPilot AI using:
- Python and Streamlit for application development.
- Google Gemini API for AI-powered financial analysis.
- PayPal Developer Platform and Sandbox REST APIs for payment integration.
- SQLite for storing invoice and transaction records.
- Pandas and Plotly for data processing and visualisation.
The application is hosted on Streamlit Community Cloud, and its source code is publicly available on GitHub.
Challenges we ran into
One of our main challenges was integrating AI-generated insights with invoice data while maintaining consistent financial calculations.
We also worked through PayPal Sandbox authentication, payment order creation, checkout approval, payment capture, transaction history, and hosted checkout redirects during cloud deployment.
Accomplishments that we're proud of
- Built and deployed a working AI-powered financial application.
- Integrated Google Gemini for invoice analysis and collection recommendations.
- Implemented PayPal Sandbox payment order creation and capture.
- Developed invoice management and payment history features.
- Published the source code in a public GitHub repository.
What we learned
This project gave us practical experience combining AI with financial applications. We learned about API integration, OAuth 2.0 authentication, payment workflows, Python application development, database management, debugging, and cloud deployment.
What's next for PayPilot AI
Our future plans include cash-flow forecasting, automated payment reminders, improved financial analytics, and more advanced AI-powered business recommendations.