Understanding Verification Results
Learn how to interpret Owl Eyes verification results, scores, and extracted data to make informed decisions about identity verification.
Verification Status
Every verification returns one of these statuses:
| Status | Meaning | Recommended Action |
|---|---|---|
| ✓ Verified | All checks passed with high confidence | Safe to proceed, identity confirmed |
| ! Review Required | Some checks flagged for manual review | Review the flags before approving |
| ✗ Failed | Verification could not be completed | Do not proceed, potential fraud |
| ○ Expired | Link expired before completion | Send a new verification link |
| ◐ Processing | Verification is being analyzed | Wait for completion (usually seconds) |
Verification Score
The verification score is a number from 0-100 representing our overall confidence in the identity verification. It's calculated from multiple weighted factors.
Score Ranges
| Score | Confidence Level | Default Status |
|---|---|---|
| 80-100 | High confidence | Verified |
| 50-79 | Medium confidence | Review Required |
| 0-49 | Low confidence | Failed |
Individual Check Results
Each verification includes multiple individual checks. Understanding these helps you make informed decisions when reviewing flagged verifications.
Document Checks
| Check | Description |
|---|---|
| Document Authenticity | Verifies security features, holograms, and document structure |
| Data Extraction (OCR) | Reads and validates text fields on the document |
| Expiration Check | Confirms the document is not expired |
| Tampering Detection | Checks for signs of digital or physical manipulation |
| MRZ Validation | Validates machine-readable zone data (passports) |
Biometric Checks
| Check | Description |
|---|---|
| Face Match | Compares selfie to document photo (similarity score) |
| Liveness Detection | Confirms a live person (not photo/video/mask) |
| Deepfake Detection | Identifies AI-generated or manipulated imagery |
Extracted Data
Owl Eyes extracts structured data from the document for your records. Here's an example of the extracted data you'll receive:
Fraud Flags
When suspicious patterns are detected, Owl Eyes raises specific fraud flags. These provide context for why a verification may have failed or requires review.
| Flag | Description | Severity |
|---|---|---|
DOCUMENT_EXPIRED | Document expiration date has passed | Medium |
FACE_MISMATCH | Selfie doesn't match document photo | High |
LIVENESS_FAILED | Failed liveness detection | High |
DOCUMENT_TAMPERING | Signs of document manipulation | High |
KNOWN_FRAUD_PATTERN | Matches known fraud pattern in threat intelligence database | High |
LOW_IMAGE_QUALITY | Image quality too low for confident analysis | Medium |
API Response Format
When receiving results via webhook or API, here's the complete response structure: