CertiTrust - Document Authentication Platform
Link to open source: https://github.com/Anshulkaocde123/CertiTrust-anshul-fork
Link to Live Project: https://web-sigma-six-19.vercel.app/
CertiTrust is a blockchain-inspired document verification system that allows institutions (universities, companies, government) to issue tamper-proof digital certificates and credentials.
Core Features
| Feature | Description |
|---|---|
| 🔐 Ed25519 Signing | Documents are cryptographically signed with institution private keys |
| 📄 PDF Stamping | QR codes embedded in PDFs contain verification data |
| 🔍 Instant Verification | Anyone can verify a document by uploading the stamped PDF |
| 🧠 AI Detection | 3-tier forensic pipeline detects manipulated/AI-generated content |
| 📊 Trust Score | Weighted algorithm (0-100%) combining crypto + forensic analysis |
| 🏛️ Multi-tenant | Multiple institutions with separate API keys and key pairs |
Tech Stack
| Layer | Technology |
|---|---|
| Frontend | Next.js 15, Tailwind CSS, shadcn/ui |
| Backend | FastAPI (Python), Gunicorn + Uvicorn |
| Database | Supabase (PostgreSQL) |
| Crypto | Ed25519 signatures, SHA-256 hashing |
| AI/ML | Hugging Face model for image manipulation detection |
| Deployment | Vercel (frontend), Render/Docker (backend) |
1. Institution uploads PDF → Backend signs with Ed25519
2. QR code with signature embedded → Stamped PDF returned
3. User uploads stamped PDF → QR decoded, signature verified
4. Forensic analysis runs → Trust Score calculated
5. Result: Valid/Invalid + Grade (A-F)
This build was uploaded as a hackathon project
