ID-Capture & Document-Authentication Web App for Fraud Prevention

Overview

A mobile-first web application that captures a customer’s ID document and selfie, validates image quality, and runs them through a commercial identity-verification service to authenticate documents and match faces, the core capture surface of a fraud-prevention tool.

The Challenge

Catching identity fraud at the point of sale means reliably capturing a usable photo of an ID and a selfie on a phone browser, rejecting blurry or glare-affected images up front, and then verifying the document’s authenticity and that the person matches it, all without a native app install.

What We Built

A React + Redux single-page app (CRA) driven by a screen flow: camera capture (a vendor camera component, CapturePhoto, CaptureSelfie), real-time image-quality gating (a dedicated sharpness/glare module plus metric toggles), Processing/Loading states, and result views for ID/passport and medical-insurance cards (Results/IDPassport, Results/MedicalCard). State is managed with redux, redux-thunk, redux-persist and connected-react-router, split into feature reducers (captured image, config, ID properties, verification, processed data). A services/api layer integrates a third-party document-authentication and facial-recognition-match platform, document verification, face match and medical-card scanning endpoints, over axios with axios-retry for resilience. The app is configured entirely through environment variables (toggles for face-match, ID+passport, medical card, mobile-only mode, sharpness/glare) and detects device type with react-device-detect, enforcing a mobile-only experience.

Technologies & Approach

React/Redux provided the predictable state needed for a multi-step capture-and-verify flow. Rather than reinventing identity verification, the app integrates a proven commercial ID-authentication/face-match SDK and focuses engineering on capture UX, on-device image-quality validation, resilient API calls (retry), and a highly configurable feature-flag-driven build.

Outcome / Impact

Delivered a configurable, mobile-web ID-capture and document-authentication front end for fraud prevention, iterated over roughly two years, the base from which a client-specific variant was later derived.

Capabilities Demonstrated

• Browser-based mobile ID and selfie capture
• On-device image quality validation (sharpness/glare)
• Integration with commercial document-authentication & face-match APIs
• Feature-flag-driven, environment-configurable React/Redux app
• Resilient API integration (axios-retry)