FaceAuthPay

Inspiration

Retail queues are slow, especially for elderly, illiterate, or first-time digital users who struggle with QR scanning and complex payment flows. Typing phone numbers or navigating apps creates friction and delays. We wanted a payment system that is fast, inclusive, contactless, and secure — something anyone can use instantly.

What it does

FaceAuthPay enables 3-second biometric payments using only a phone camera. Users simply look at the screen, pass a 3D liveness check, and get automatically redirected to their GPay/PhonePe/Paytm for final confirmation. No QR codes, no typing, no confusion. It uses encrypted neural face embeddings (not photos) and zero-knowledge storage to ensure privacy and security.

How we built it

Neural face embedding model (128-D vector representation)

3D liveness detection to prevent spoofing

Client-side AES-256 encryption before sending any data

OTP + Face 2FA for high-value transactions

JavaScript + HTML + CSS + Base44 for the frontend workflow

API for linking embeddings to UPI apps and redirecting payments

Challenges we ran into

Ensuring accuracy while keeping the verification under 3 seconds

Building strong spoof detection using only mobile cameras

Making biometric storage privacy-safe with zero-knowledge design

Handling users without QR codes or saved numbers

Accomplishments that we're proud of

Achieved sub-3-second verification and payment redirection

Built QR-less payments using face-only identification

End-to-end encrypted biometric pipeline

Designed a system elderly users can operate instantly

What we learned

We learned how difficult real-time liveness detection is on low-end devices, how to build privacy-first biometric systems, and how small UX changes dramatically reduce checkout time. We also realized how important inclusivity is in payments.

What’s next for FaceAuthPay

Integration with offline stores and self-checkout kiosks

Support for multi-user households and shared devices

Face-only payments without opening any app

On-device embedding matching for even stronger privacy

Expansion to metro gates, tolls, and public transport