Skip to main content
Open-source · Self-hosted · Event-driven

AI-powered trading,
orchestrated by events.

AITradingSystem connects market triggers to multi-agent AI trading groups, producing reasoned decisions and executing orders — all on your own infrastructure.

AITradingSystem — quick deploy

$ mkdir AITradingSystem && cd AITradingSystem

$ wget https://raw.githubusercontent.com/QuantaGenesis/AITradingSystem/main/install.sh

$ chmod +x install.sh

$ ./install.sh setup

$ ./install.sh start

✓ backend ready on :8000

✓ frontend ready on :3333 (or custom port)

✓ database healthy

How it works

Four stages, one coherent event flow. Each component is a standalone service that communicates over a secure WebSocket gateway.

1
Trigger Service
Monitors news, market data, or social signals
2
Event Gateway
Validates, persists, and routes events via WebSocket
3
Trading Group
AI agents discuss and produce a trade decision
4
Exec Engine
Places orders on the configured exchange

Everything you need

Event-Driven Core

Market signals flow from triggers through the WebSocket gateway to trading groups in real time. No message queue required — FastAPI handles routing in-process.

🤖

Multi-Agent Discussion

Each trading group assembles a team of AI agents — analyst, strategist, risk manager — that debate every signal before the group leader makes a decision.

🔌

Pluggable Triggers

Copy the analyzer template, replace the data source, and your custom trigger (news, RSI, social sentiment) connects to the same gateway with zero backend changes.

📊

Exchange Execution

Binance and OANDA adapters execute decisions. Simulated → Sandbox → Live rollout gives you full confidence before touching real capital.

🧠

Experience System

Every closed trade is analyzed by an LLM extraction pipeline and stored as a searchable lesson. Future discussions retrieve relevant past experiences automatically.

🔒

Secure by Design

Single-session token enforcement, role-scoped WebSocket permissions, startup secret guards, and security headers ship out of the box.

Ready to deploy?

One Docker Compose command. Your own infrastructure. Full control.