PayWiser

For merchants, PayWiser eliminates costly transaction fees (2–3% in legacy systems) and settlement delays (often 1–3 days). Funds arrive instantly, improving cash flow and reducing dependency on intermediaries. For consumers, PayWiser offers unmatched convenience—no wallets, phones, or PINs required—while ensuring privacy through encrypted, non-custodial identity verification.

Security enhancements include Bluetooth proximity verification, ensuring the user’s device is within 1 metre of the merchant’s terminal to prevent fraud. Additionally, optional voice recognition provides a secondary authentication layer for added security.

By combining biometric authentication with blockchain-based settlement and advanced proximity and voice verification, PayWiser creates a payment experience that is faster, cheaper, safer, and more inclusive than legacy systems.

PayWiser represents a sophisticated integration of cutting-edge biometric authentication with decentralized finance infrastructure. At its core, we built a multi-layered architecture combining FastAPI-powered DeepFace recognition services with Node.js/Express APIs, all orchestrated through a comprehensive smart contract ecosystem. The face recognition system leverages OpenCV and DeepFace libraries to generate facial embeddings, which are securely stored using SAGA's distributed storage network - this partnership proved crucial for maintaining both data privacy and accessibility across our decentralized infrastructure. The biometric data never touches centralized servers, instead being cryptographically hashed and stored on SAGA's blockchain-native storage layer, ensuring user privacy while enabling seamless cross-chain authentication.

The technical backbone integrates Circle's Programmable Wallets SDK for seamless USDC transactions with Hyperlane's interoperability protocol for cross-chain settlements, creating a truly universal payment experience. Our smart contracts, built with Hardhat and deployed across multiple testnets including Sepolia, handle payment escrow and settlement logic while interfacing directly with Hyperlane's message passing system. The particularly innovative aspect lies in our custom Hardhat tasks that bridge biometric verification with on-chain execution - when a user's face is verified through our DeepFace API, it triggers cryptographic proofs that unlock smart contract functions, essentially turning facial recognition into a decentralized key management system. This eliminates the need for traditional private key handling while maintaining the security guarantees of blockchain transactions.

What makes this implementation particularly noteworthy is our "hacky" but elegant solution to the oracle problem for biometric data. Instead of relying on centralized oracles, we created a hybrid verification system where facial embeddings stored on SAGA are cross-referenced with real-time captures, generating zero-knowledge proofs that confirm identity without exposing the actual biometric data. The most technically challenging aspect was synchronizing the face verification workflow with Hyperlane's cross-chain messaging - we implemented a custom event-driven architecture where successful face authentication on one chain triggers settlement transactions on another, all while maintaining sub-second response times. This required building sophisticated retry mechanisms and state management across multiple blockchain networks, ultimately creating a payment system that feels as seamless as traditional fintech while leveraging the full security and decentralization benefits of Web3 infrastructure.