A platform for conducting financial market experiments with multiple traders and role-based interactions.
- Lightweight session pools with lazy market creation
- Zero resource waste from zombie markets
- Informed traders (buy/sell specialists)
- Speculators (flexible strategies)
- Automatic goal-based trader matching
- Coordinated market starts when all traders ready
- Real-time status updates and waiting room
- Seamless session-to-market conversion
- Live order book and price charts
- Instant execution feedback
- Market summary with P&L and VWAP
- Google authentication
- WebSocket communication
- Comprehensive error handling
- Modular event-driven architecture
Session pool mechanism handles multiple human traders elegantly:
- Users join lightweight session pools (not heavy markets)
- Goal assignment based on predefined slots (e.g., [100, -200, 150])
- Real-time trader count updates ("2 out of 3 traders joined")
- All traders ready → automatic market creation
- Heavy infrastructure created only when needed
- Human + AI traders added, trading begins
- Fast user onboarding
- Zero resource waste
- Automatic role assignment
- Seamless scaling
bash <(curl -sSL https://raw.githubusercontent.com/dthinkr/trading_platform/main/trading_platform_run.sh)# Frontend
cd front
npm install
npm run dev
# Backend
cd back
pip install -r requirements.txt
uvicorn api.endpoints:app --reloadFor detailed documentation, feature explanations, and API references, please visit our Wiki.
See our Changelog for detailed version history and updates.
MIT License - see LICENSE for details.
