Build a production-ready trust infrastructure layer that verifies a real human is behind a digital interaction — designed for social platforms — in a world flooded with deepfakes and voice clones.
Core vision:
As deepfakes and AI-generated voices/faces proliferate, we need a way to prove there's an authentic human on the other side — without sacrificing privacy. Build the verification layer that apps in social platforms can drop in to establish "this is a real, unique human."
The system must deliver a complete verification loop:
1. Let a user complete a human-verification flow
2. Issue a privacy-preserving proof of humanity (not raw biometrics)
3. Let apps request and verify that proof
4. Prevent fraud, bots, and impersonation
5. Preserve user privacy and minimize data retention
Core capabilities:
Verification flow (user-facing):
- A smooth, quick check that confirms liveness and humanity
- Privacy-first: issue a reusable proof/credential, not stored raw biometrics (represented at the app layer)
- Clear consent and explanation of what's checked and kept
- Re-verification and revocation controls for the user
Developer platform:
- API key issuance and a copy-paste integration snippet
- Endpoints to request verification and check a user's proof
- Webhooks for verification results
- Sandbox mode for testing
Fraud prevention:
- Uniqueness checks to stop duplicate/bot accounts
- Signals and risk scoring for suspicious attempts
- Rate limiting and abuse protection
- A verification event log per app
Privacy and trust:
- Data minimization by default; no unnecessary retention
- Transparent privacy model and user-facing controls
- Auditable records without exposing sensitive data
Design:
Clean, secure, and reassuring — a trust-forward aesthetic with clear consent UX for users and a crisp, developer-friendly dashboard for integrators.
Builds a working MVP of a human-verification trust layer. Includes a verification flow that issues a privacy-preserving proof of humanity, a developer-facing dashboard with an API key and integration snippet, and a verification event log. Biometric capture and cryptographic attestation are represented at the application layer.