AEGIS Trading System

Automated trading stack combining FinRL (PPO), multi-agent analysis (news, social, market, SEC), and real-time market data over Redis and MongoDB. A Next.js dashboard talks to a unified FastAPI backend.

Documentation

Topic	Guide
Install & configure	docs/setup.md
Run the system	docs/quickstart.md
Architecture	docs/architecture.md
All topics (index)	docs/README.md

Prerequisites

• Python 3.11+
• Redis 7+, MongoDB 7+
• API keys: Alpaca (paper), OpenAI; optional NewsAPI / SEC

See docs/setup.md for full requirements and .env variables.

Quick start

From the repository root (with .env configured):

pip install -r requirements.txt

# Optional: historical bars in MongoDB (first-time)
python historical_data.py

# Trading pipeline (recommended)
python -m backend.src.cli --continuous

# API + auth (separate terminal)
python main.py

# Frontend (separate terminal)
cd frontend && pnpm install && pnpm dev

Fast local tests:

./scripts/ci/run-tests.sh

Details: docs/quickstart.md · docs/testing.md

Docker

docker compose up -d

Services: Redis, MongoDB, aegis-pipeline, aegis-api (port 8000), aegis-frontend (port 3000). See docs/docker.md.

Repository layout

Hedgera/
├── backend/          # Python: API, agents, FinRL, orchestration
├── frontend/         # Next.js dashboard
├── docs/             # Detailed documentation
├── scripts/ci/       # Local test runner (no GitHub Actions yet)
├── main.py           # Starts FastAPI (backend.src.api)
└── requirements.txt  # Delegates to backend/requirements.txt

Full tree: docs/project-structure.md

License

See repository license file if present.