Skip to main content

Fraud Detection Overview

The Fraud Detection Service by Veryfi supplements the core data extraction API, designed to identify potential fraudulent documents submitted for processing. This service scrutinizes each submission using sophisticated detection methods and flags suspicious content directly in the JSON response alongside your extracted data.

Two-Stage Fraud Detection Approach​

Veryfi's Fraud Detection employs a comprehensive 2-stage approach that combines mobile capture technology with AI-powered cloud processing to minimize fraud risks:

  1. Initial Screening (Veryfi Lens): Performs preliminary device security checks, image quality validation, and source verification during document capture
  2. Advanced Analysis (Cloud AI): Applies computer vision, natural language processing, and behavioral pattern recognition to detect sophisticated fraud attempts
Fraud Detection Flow Diagram

Detection Capabilities​

The fraud detection system utilizes Signals to analyze document authenticity and employs Triggers to flag suspicious patterns in the JSON response. These signals include:

  • Device security validation
  • Image quality assessment
  • Content verification
  • Pattern recognition
  • History matching
  • Velocity tracking (requires Veryfi Lens)

Integration​

Fraud detection results are conveniently integrated into the same API response you use for data extraction, eliminating the need for separate fraud checking calls. This streamlined approach ensures you receive both extracted data and fraud analysis in a single JSON response.

For enhanced detection capabilities and velocity tracking features, Veryfi Lens mobile capture is recommended as an additional component of your fraud prevention strategy.