A decentralized prediction markets platform powered by AI oracles on the Sui blockchain.

Live Demo: aletheia-app.vercel.app

Aletheia (Greek for "truth") is a next-generation prediction market platform that leverages AI-powered oracles and blockchain technology to create trustless markets for forecasting real-world events. Built on the Sui blockchain, it enables users to trade on the outcome of events with automated market making and AI-driven resolution.

• Framework: Next.js 14 with TypeScript
• Styling: Tailwind CSS
• Charts: Chart.js, Recharts
• Web3: Sui dApp Kit

• Runtime: Node.js with Express
• Database: MongoDB with Mongoose
• Real-time: Socket.io

• Platform: Sui
• Language: Move
• Features: Automated Market Making (LMSR), SEAL encryption

• AI Models: Multiple LLM providers for consensus
• Resolution: Automated market resolution with multi-model validation

• Multi-outcome markets: Support for 2-10 possible outcomes
• Automated Market Making: LMSR (Logarithmic Market Scoring Rule) algorithm
• Real-time odds: Live probability updates based on trading activity
• Liquidity provision: Users can add liquidity to markets

• Privacy: SEAL encryption for private markets
• Time-locked: Markets can be sealed until a specific date
• Premium features: Encrypted comments and exclusive access

• Multi-model consensus: Uses multiple AI models for resolution
• Automated resolution: Markets resolve automatically based on real-world outcomes
• Nautilus integration: External data verification

• Live probability charts: Visualize market sentiment over time
• Portfolio tracking: Track your positions and P&L
• Market history: Historical data and trends

Aletheia/
├── backend/           # Express.js API server
│   ├── src/
│   ├── package.json
│   └── ...
├── frontend/          # Next.js frontend
│   ├── src/
│   ├── package.json
│   └── ...
├── oracle/            # AI oracle service
│   ├── src/
│   ├── package.json
│   └── ...
├── contracts/         # Sui Move smart contracts
│   └── sources/
├── e2e/               # End-to-end tests
│   ├── package.json
│   └── ...
└── package.json       # Root workspace config

npm run devs              # Build all and run all dev servers
npm run build:all         # Build backend, oracle, and frontend
npm run dev:all           # Run all services in parallel
npm run dev:backend       # Run backend on port 3001
npm run dev:oracle        # Run oracle service
npm run dev:frontend      # Run frontend on port 3000
npm run test              # Run E2E tests

• Node.js 18+
• MongoDB instance
• Sui wallet with testnet/mainnet SUI
• API keys for AI providers

# Install all dependencies
npm install

# Or install individually
cd backend && npm install
cd frontend && npm install
cd oracle && npm install

Create .env files in each directory (see .env.example files for reference).

# Run all services in development mode
npm run devs

# Or run individually
npm run dev:backend   # Backend on port 3001
npm run dev:oracle    # Oracle service
npm run dev:frontend  # Frontend on port 3000

# Build all services
npm run build:all

# Run E2E tests
npm test

1. Market Creation: Anyone can create a prediction market with multiple outcomes
2. Trading: Users buy shares in outcomes they believe will occur
3. Price Discovery: LMSR algorithm adjusts prices based on demand
4. Resolution: AI oracles determine the outcome after market close
5. Settlement: Winners receive payouts proportional to their shares

The platform uses several Move modules:

• prediction_market: Core market functionality
• amm: Automated market making
• oracle: Oracle integration
• seal: Encryption for sealed markets

• Setup Instructions
• Deployment Guide
• Sealed Markets Guide

• Mainnet deployment
• Mobile app
• Advanced charting
• Social features (following, profiles)
• Token incentives
• Cross-chain markets

Preyam - GitHub

MIT