NoaSight
The Future of Prediction Markets
About
The problem it solves NoaSight is a decentralized prediction market platform built on the BASE blockchain, inspired by platforms like Polymarket but leveraging BASE's superior transaction speed and cost efficiency. - Lightning-fast transactions on BASE testnet - Unique NFT profile system with benefits - Community-driven oracle resolution - DAO governance for platform decisions
How it's made
Challenges we ran into During the development of NoaSight, the biggest challenge was bridging the gap between Web2 usability and Web3 complexity, specifically regarding Wallet Abstraction and Cross-Chain Synchronization. 1. The "Gas Fee Barrier" & Seamless Onboarding The Hurdle: We wanted users to trade prediction shares instantly using just their email, without worrying about seed phrases or gas fees (ETH/SUI). However, on-chain transactions inherently require gas, which is a major friction point for non-crypto natives. Integrating standard wallets meant users had to buy ETH first, killing the conversion rate. The Solution: We implemented a robust Gas Sponsorship System combined with Circle Programmable Wallets. - I built a backend service (treasurySweeper.js and gas monitors) that monitors user wallets created via Circle. - When a user initiates a trade (buy/sell), our backend verifies the transaction intent and automatically sponsors the gas fees using a centralized Treasury Wallet on Base Sepolia. - This creates a truly "gasless" experience where users just see their balance update, while complex blockchain interactions happen invisibly in the background. 2. Synchronizing Off-Chain & On-Chain State The Hurdle: A prediction market requires real-time speed. Relying solely on blockchain confirmations made the UI feel sluggish. However, relying on a database introduced the risk of "state drift" where the user's displayed balance didn't match their actual on-chain assets, especially after failed transactions or reorgs. The Solution: I architected a Dual-State System with Self-Healing Monitors: - Optimistic UI: The frontend updates instantly based on API responses. - Background Monitors: I created specialized background services (suiDepositMonitor.js, circleDepositMonitor.js) running on the backend. These services constantly poll the blockchain RPCs to detect deposits and confirm transaction finality. - Auto-Retry & Sweep: If a minting or transfer transaction fails (e.g., due to RPC timeout), the system automatically retries with exponential backoff. A "Treasury Sweeper" service ensures that funds deposited by users are securely aggregated to the main liquidity pool periodically, keeping the ecosystem solvent and liquid. 3. Multi-Chain Compatibility (EVM + Sui + Solana) The Hurdle: Designing a backend that could handle prediction markets across different architectures (EVM-based Base Chain vs. Move-based Sui vs. Solana) was architecturally demanding due to different address formats, signing mechanisms, and RPC behaviors. The Solution: I adopted a Modular Service Architecture. Instead of a monolithic codebase, I separated blockchain logic into distinct service modules (suiWalletService.js, solanaIntegration.js, standard EVM providers). - This allowed me to abstract the logic: The core market engine doesn't care which chain a user is on; it just calls a standardized createWallet() or getBalance() interface. - I used ethers.js for EVM chains and dedicated libraries/APIs for Sui/Solana, unifying them under a single PostgreSQL schema that maps generic wallet_id and derivation_index to specific chain implementations.