Build, test, and evolve trading strategies with AI — entirely on your own machine. No cloud subscription. No data leaks. No monthly fees.
TradeSight is a self-hosted Python app that runs AI-powered strategy tournaments overnight, backtests technical indicators, and executes paper trades via Alpaca — all from a local web dashboard.
- Algorithmic trading hobbyists who want to test strategies without risking real money
- Python developers exploring quantitative finance and AI-driven decision systems
- Privacy-conscious traders who don't want their strategies on someone else's server
- Makers building autonomous financial agents
| Feature | Description |
|---|---|
| 🧬 AI Strategy Tournaments | Automated overnight evolution of trading strategies — the best wins, rest are retired |
| 📊 15+ Technical Indicators | MACD, RSI, Bollinger Bands, EMA crossovers, ATR, volume analysis, and more |
| 💸 Paper Trading | Connect Alpaca paper account — trade with fake money, track real P&L |
| 🔍 Multi-Market Scanner | Scan stocks + Polymarket prediction markets for signals simultaneously |
| 🌐 Web Dashboard | Real-time Flask interface — positions, signals, tournament results, logs |
| ⏰ Cron Automation | Overnight strategy improvement runs automatically — wake up to new results |
| 🔒 100% Local | Runs on your machine. Your strategies stay yours. |
- Python 3.11+
- macOS or Linux (Windows via WSL)
- Alpaca paper trading account (free, optional — demo mode works without it)
macOS (Homebrew):
brew tap rmbell09-lang/tradesight
brew install tradesightFrom source:
git clone https://github.com/rmbell09-lang/tradesight.git
cd tradesight
pip install -r requirements.txtpython START_TRADESIGHT.pyDashboard opens at http://localhost:5000
TradeSight runs fully in demo mode with simulated market data — no Alpaca account needed to explore.
- Create a free Alpaca paper account
- Add your API keys to
config/api_keys.json:
{
"alpaca_key": "YOUR_KEY",
"alpaca_secret": "YOUR_SECRET",
"paper": true
}┌─────────────────────────────────────────────────┐
│ TradeSight Dashboard [Localhost:5000] │
├──────────┬──────────┬───────────┬───────────────┤
│ Markets │Tournaments│ Trading │ Settings │
├──────────┴──────────┴───────────┴───────────────┤
│ Active Signals: 3 Open Positions: 2 │
│ Best Strategy: MACD Crossover (score: 0.72) │
│ Paper P&L: -$113.96 (initial RSI strategy) │
│ Next Tournament: Tonight @ 2:00 AM │
└─────────────────────────────────────────────────┘
169/169 tests passing ✅
python -m pytest tests/ -vtradesight/
├── src/
│ ├── scanner.py # Multi-market signal scanner
│ ├── strategy_lab/ # AI tournament engine
│ ├── trading/ # Alpaca paper trade executor
│ ├── indicators/ # 15+ technical indicators
│ └── automation/ # Overnight cron jobs
├── web/ # Flask dashboard
├── config/ # API keys + settings
├── data/ # Price history cache
└── tests/ # 169 unit tests
Real money-at-risk (on paper). Updated automatically. System has been live since March 2026.
Current Portfolio: $532 total value | +6.43% total return (+$32.15)
| Strategy | Trades | P&L | Win Rate |
|---|---|---|---|
| MACD Crossover | 1 | +$17.89 | 100% ✅ |
| VWAP Reversion | 1 | -$4.75 | 0% |
| RSI Mean Reversion | 5 | -$13.77 | 20% |
Recent Closed Trades (Last 7 Days):
| Symbol | Strategy | Entry | Exit | P&L |
|---|---|---|---|---|
| JPM | MACD Crossover | $242.94 | $303.24 | +$17.89 ✅ |
| SPY | RSI Mean Reversion | $596.36 | $649.79 | +$6.06 ✅ |
| AMZN | VWAP Reversion | $223.25 | $209.69 | -$4.75 |
| QQQ | RSI Mean Reversion | $590.06 | $565.22 | -$6.56 |
| GOOG | RSI Mean Reversion | $297.77 | $276.47 | -$13.28 |
This is a live paper trading account connected to Alpaca. All prices are real market data; no real money at risk. System runs overnight tournaments to evolve strategy parameters automatically. Early RSI underperformance drove a regime-filter upgrade now in production.
- Multi-indicator technical analysis (15+ indicators)
- AI strategy tournament engine
- Alpaca paper trading integration
- Real-time web dashboard
- Overnight automation (cron)
- Phase 1: Active stop-loss + take-profit execution
- Phase 1: Trailing stop with high-water mark
- Phase 2: Confluence strategy (multi-indicator entry gates)
- Phase 2: Market regime detection (bull/bear/sideways filter)
- Phase 3: Monte Carlo simulation for strategy validation
TradeSight is MIT-licensed and free to use. If it saved you time or you want the packaged strategy lab with setup guide and pre-tuned parameters:
Get TradeSight Strategy Lab on Gumroad → — $49 one-time
Includes: packaged download, setup walkthrough, pre-configured Alpaca integration, and strategy parameter reference.
TradeSight is similar to — but different from — these popular Python trading tools:
| Project | What It Does | How TradeSight Differs |
|---|---|---|
| backtrader | Python backtesting framework | TradeSight adds AI strategy evolution + live paper trading web dashboard |
| freqtrade | Crypto trading bot | TradeSight focuses on stocks (Alpaca API) with overnight strategy tournaments |
| vectorbt | Vectorized backtesting in notebooks | TradeSight is a self-hosted web app — no notebook required |
| zipline-reloaded | Algorithmic trading library | TradeSight is a full app, not a library — no Python trading experience needed |
| Jesse | Crypto strategy framework | TradeSight is for stocks + prediction markets, not crypto |
| nautilus_trader | High-performance trading platform | TradeSight is simpler, self-hosted, built for hobbyists not HFT |
| BillingWatch | Self-hosted billing anomaly detection | Same maker — catch Stripe issues before they cost you |
Also useful for: python trading bot · algorithmic trading python · paper trading software · free backtesting · stock trading python · quantitative finance · algo trading strategy tester · self-hosted trading platform · automated trading system · backtesting framework
MIT — free to use, modify, and build on.
Star the repo — it helps other Python traders find it.
Got broken AI-generated code? → Vibe Code Rescue