KYC Identity Verification

To build this KYC identity verification app, I leveraged my expertise in Python (FastAPI) to create a high-performance, asynchronous backend capable of handling multiple verification requests simultaneously. I integrated advanced Computer Vision technologies, utilizing OpenCV, MediaPipe for liveness detection, and InsightFace for precise biometric matching, alongside Anthropic's Claude Vision API for intelligent OCR document extraction. On the frontend, I used modern JavaScript with native HTML5 APIs to manage real-time camera streams and video recording directly in the browser. I also implemented Redis for efficient session management and caching, ensuring the system remains responsive and scalable under high concurrency. My focus was on creating a secure, non-blocking architecture that safely processes sensitive user data.

This application serves as a robust Identity Verification System, designed to securely onboard users by validating their physical identity against their official documents in real-time. It solves the critical problem of fraud prevention and regulatory compliance (KYC/AML) for digital platforms. Users simply scan their ID card or passport and take a selfie video; my system then automatically extracts their data, verifies the document's authenticity, matches their face biometrics to the ID photo, and performs a "liveness check" to ensure a real person is present, preventing spoofing attacks with photos or screens. It streamlines what is typically a manual, slow process into a fast, automated, and secure digital experience.