Didit's Biometric Authentication solution provides a streamlined verification experience for returning users. This workflow uses facial recognition to confirm identity without requiring document re-verification, creating a frictionless experience while maintaining high security standards.

Key Features
Fast Re-Verification
- No document scanning required
- Complete verification in seconds
- Reduces user friction and abandonment
Advanced Security
- Uses the same neural network architecture as Face Match 1:1
- Prevents account takeover attempts
- Includes liveness detection to prevent spoofing
Integration Flexibility
- Available as web-based
- Configurable matching thresholds
- Optional IP analysis for enhanced security
How It Works
-
When you create a biometric authentication session:Session Creation- You must provide the
portrait_image
in Base64, which can be extracted from previous verification sessions, or just other images you already have saved in your DB from the user. - The biometric authentication workflow is initialized
- You must provide the
{
// Replace with biometric authentication workflow id
"workflow_id": "11111111-2222-3333-4444-555555555555",
"vendor_data": "user-123",
"callback": "https://example.com/verification/callback",
"metadata": {
"login_attempt": "2"
},
"portrait_image": "/9j/4AAQSkZJRgABAQEAyQDJAAD/2...Y+QrTcpH/9k="
}
During the authentication process:
- User takes a new selfie (with active liveness detection)
- System evaluates image quality
- Poor quality images are rejected with guidance for improvement
- User can retry if needed
-
The system performs a sophisticated face matching process:Biometric Comparison- Compares new selfie with `portrait_image
- Analyzes facial landmarks and features
- Generates similarity score (0-100%)
- Applies configured threshold for decision
-
Based on the comparison results:Result Processing- Scores above threshold → Authentication approved
- Scores below threshold → Authentication declined
- Results are available via API, dashboard, and webhooks