From Arduino optical sensors to Mantra MFS110 enterprise hardware — a full-stack TypeScript biometric voting infrastructure with zero-knowledge proofs and Merkle-tree audit trails.
Two-factor auth with credentials + ISO fingerprint scan via AES-256-GCM pipeline.
Click to deep dive →SHA-256 chained vote blocks — any alteration breaks the entire chain above it.
Click to deep dive →Real-time queue health, anomaly detection, VVPAT reconciliation via React dashboard.
Click to deep dive →C++ Arduino → TypeScript + PostgreSQL. R307 sensor → Mantra MFS110 UIDAI device.
Click to deep dive →Biometric template encrypted on-device. Nothing traverses the wire in cleartext.
Click to deep dive →Digital votes mirrored to physical paper slips for forensic audit reconciliation.
Click to deep dive →Cipher Suite Active — No Plaintext
ISO 19794-2 Template Matching
Distributed Ledger Consensus
Duplicate template rejections
Successful biometric authentications
Mantra polling — Port 11100
Basic optical fingerprint sensor via UART. No encryption. Plaintext EEPROM storage.
Replaced R307 with UIDAI-certified hardware. Introduced HTTP RD Service layer and ISO 19794-2 templates.
C++ → Node.js/Express + React 19. PostgreSQL with parameterized queries and JWT auth.
Zero-knowledge biometric pipeline. Raw templates never stored. JWT + RBAC guards on all routes.
SHA-256 vote chaining, VVPAT paper reconciliation, admin dashboards, and multi-role access control.
Full grid control — election lifecycle, hardware oversight, fraud triage.
The full Mantra MFS110 mock acquisition flow — ISO-template extraction, AES-256 encryption, PostgreSQL hash lookup, and Merkle-tree block append.
:: AES-256-GCM VAULT SEALED ::