Inspiration

Online payments often require users to navigate multiple fields and manually enter transaction details. We wanted to make this process simpler and more intuitive using AI. This inspired us to build PayPilot AI, an application that converts natural-language payment requests into structured PayPal transactions.

What it does

PayPilot AI allows users to describe a payment in simple natural language. Gemini AI understands the request and extracts important details such as payment amount, currency, and description. If information is missing or unclear, it asks for clarification instead of making assumptions.

Once the details are ready, users can review the payment information and proceed through PayPal Sandbox checkout. The application also maintains transaction history so users can track their payment activity.

How we built it

We built PayPilot AI using React, TypeScript, and Vite for the frontend, with Python and FastAPI powering the backend.

We integrated Google Gemini API to understand natural-language payment requests and convert them into structured payment details. PayPal Orders API v2 handles order creation and payment capture, while the PayPal JavaScript SDK provides the checkout experience.

We also used SQLite to maintain transaction records and implemented backend validation and error handling to improve reliability. PayPal Sandbox was used for payment integration and testing.

Challenges we ran into

One of our main challenges was accurately extracting payment details from natural-language requests. Users may provide incomplete information or express the same payment intention in different ways.

We addressed this by integrating Gemini AI with structured response validation and clarification handling. Another challenge was integrating PayPal Sandbox checkout with the frontend and backend while managing API responses and payment states.

We also focused on protecting sensitive API credentials and ensuring that AI-generated details are reviewed by users before proceeding with a payment.

Accomplishments that we're proud of

We are proud to have built a working AI-powered payment application that connects natural-language understanding with PayPal payment infrastructure.

Our key accomplishments include:

  • Integrating Gemini AI for natural-language payment interpretation.
  • Extracting payment amounts, currencies, and descriptions into structured data.
  • Handling incomplete requests through clarification.
  • Integrating PayPal Orders API v2 and PayPal JavaScript SDK.
  • Implementing transaction history using SQLite.
  • Building a user-controlled checkout flow where AI never authorizes payments automatically.

What we learned

Through this project, we learned how to integrate generative AI with real payment infrastructure. We gained practical experience with Gemini API, PayPal Orders API, React, FastAPI, REST APIs, and database integration.

We also learned the importance of validating AI-generated outputs, handling incomplete user inputs, securing API credentials, and designing payment workflows that keep

Built With

Share this project:

Updates

Submission history