5-signal fusion AI — Pattern Recognition, FinBERT Sentiment, LSTM, TCN, and ML Ensemble — delivering 85.2% accuracy across 259 assets in live paper trading.
From signal generation to live order execution — a complete trading intelligence stack built on peer-reviewed research.
RF+XGBoost ensemble, chart pattern recognition, FinBERT news sentiment, LSTM, and TCN dilated causal convolutions combined with dynamic weighting.
Real-time NLP on financial news with YouTube intelligence blending. Detects high-impact catalysts and adjusts signal weights automatically.
35+ chart patterns detected in real-time including head & shoulders, double tops, triangles, flags, and wedges across all timeframes.
Compact LSTM + PCA for sequential pattern memory, plus 8-layer Temporal Convolutional Network with 256-bar receptive field for multi-scale analysis.
PPO reinforcement learning agent that adapts position size based on signal confidence, portfolio volatility, win rate, and market regime in real time.
Fully automated paper and live trading via Alpaca brokerage. Kelly criterion sizing, trailing stops, ATR-based exits, and drawdown circuit breakers.
US stocks, cryptocurrencies (BTC, ETH, SOL, XRP…), forex pairs, ETFs, and futures. Multi-class portfolio with regime-aware allocation.
Insider transaction signals from SEC Form 4 filings combined with options flow sentiment for institutional-grade equity intelligence.
Automated Sunday retraining of all models on rolling 6-month windows with accuracy benchmarking and auto-promotion when models improve.
IntelliTradeAI originated as peer-reviewed academic research at Prairie View A&M University and has been validated in live paper trading.
Peer-reviewed paper documenting the multi-signal fusion architecture, 85.2% accuracy benchmark, and walk-forward validation methodology.
Developed at Prairie View A&M University's engineering department with rigorous academic standards and reproducible experimental protocols.
All performance metrics use out-of-sample walk-forward backtesting — no look-ahead bias. Sharpe 1.74, max drawdown <12% in live paper trading.
Real-time Alpaca paper trading with verified P&L tracking. All trades, win rates, and drawdown stats are recorded and available via API.
Recent closed trades from the live Alpaca paper trading session. All results are real — no simulated or cherry-picked data.
| Symbol | Direction | Entry Signal | Outcome | R-Multiple | Close Reason |
|---|---|---|---|---|---|
| NVDA | BUY | Fusion: 91% conf | WIN | +2.1R | Take profit reached |
| BTC-USD | BUY | Fusion: 88% conf | WIN | +1.8R | Take profit reached |
| AAPL | SELL | Fusion: 84% conf | WIN | +1.4R | Take profit reached |
| ETH-USD | BUY | Fusion: 79% conf | LOSS | -0.9R | Stop loss triggered |
| MSFT | BUY | Fusion: 87% conf | WIN | +2.3R | Take profit reached |
| SOL-USD | BUY | Fusion: 83% conf | WIN | +1.6R | Take profit reached |
* Paper trading results. Past performance does not guarantee future results. All trading involves risk.
All plans include a 7-day free trial. No credit card required to start.
| Feature | Starter | Pro | Elite |
|---|---|---|---|
| Assets monitored | 50 | 259 | 259+ |
| ML + Pattern + News signals | ✓ | ✓ | ✓ |
| LSTM deep learning signal | — | ✓ | ✓ |
| TCN (5th fusion signal) | — | ✓ | ✓ |
| Live Alpaca paper trading | — | ✓ | ✓ |
| Live real-money trading | — | — | ✓ |
| RL position sizing (FW6) | — | ✓ | ✓ |
| SEC EDGAR insider signals | — | — | ✓ |
| Options flow intelligence | — | — | ✓ |
| Plaid connected accounts | — | — | ✓ |
| Weekly model retraining | — | ✓ | ✓ |
| API access | — | ✓ | ✓ |
Join traders using IEEE-published AI to analyze 259 assets in real time.
No credit card required. Cancel anytime.