Automated KYC Document Verification

I need an automated system that carries out document-based KYC verification end-to-end without manual intervention. Users will upload their files through a simple web or mobile interface; the system must instantly read the document, extract key data, detect tampering, and return a pass/fail result with confidence scores.

You are free to choose the tech stack, but accuracy and speed are critical. Leveraging OCR libraries such as Tesseract or Google Vision, fraud-detection APIs, or machine-learning models is fine as long as the final workflow is fully automated and can be audited. A clean REST or GraphQL endpoint for integration into my existing onboarding flow is required.

Deliverables
• Source code and clear build instructions
• Deployed demo (cloud or container) showing real-time verification
• API documentation with request/response examples
• A short report explaining the verification logic and measurable accuracy

Successful completion will be confirmed once the demo consistently verifies sample documents in under five seconds and the codebase passes my security review.