Inspiration
Online payment scams can happen at the moment a user is about to make a payment. We wanted to create a simple safety layer that helps users pause and understand potential warning signs before continuing to checkout.
What We Built
PayPal ScamShield is an AI-powered payment safety assistant. A user enters the merchant name, payment amount, description, and whether they know the merchant. Gemini AI analyzes the provided information and returns a risk level, score, warning reasons, and a recommendation.
After reviewing the AI analysis, the user can either cancel or continue to PayPal Sandbox checkout.
How We Built It
The application was built using Node.js and Express for the backend, with HTML, CSS, and JavaScript for the frontend.
We integrated:
- Gemini AI for payment risk analysis
- PayPal JavaScript SDK for checkout
- PayPal Orders API for creating and capturing Sandbox orders
- PayPal Sandbox for testing the payment flow
Challenges
One challenge was connecting the AI analysis flow with the PayPal checkout flow while keeping payment credentials secure. We used environment variables for API credentials and kept them out of the public repository.
Another challenge was handling AI responses reliably, so the application validates the returned risk level, score, reasons, and recommendation before displaying the result.
What We Learned
We learned how to combine an AI service with a real payment platform workflow and how to build a user-facing safety layer before a payment decision.
ScamShield is an AI risk-analysis prototype. It does not determine whether a merchant is fraudulent and does not guarantee payment safety.
Built With
- ai
- api
- css
- express.js
- gemini
- html
- javascript
- node.js
- orders
- paypal
- sandbox