Dr. Jaime M.

NinjaScript, C#, MQL5, PineScript, Python Developer | ML AI | PhD

I build trading systems that survive live execution — not just backtesting. 27 years of C# development. 20+ NinjaTrader 8 indicators and strategies. 10+ MetaTrader 5 Expert Advisors. Python ML pipelines with walk-forward validation and ONNX deployment. PhD in Management Information Systems. Every system I deliver comes with full source code, documentation, and a walkthrough so you understand what I built and why. I'm a developer who understands trading, not a trader who learned to code. That difference matters when your strategy needs to handle partial fills, slippage on a fast NQ breakout, or a prop firm's trailing drawdown rules. WHAT I BUILD NinjaTrader 8 (NinjaScript/C#): Custom indicators (regime filters, order flow, harmonic patterns, composite dashboards), automated strategies (breakout, momentum, mean-reversion, institutional zone detection, volatility compression), advanced trade management (multi-target exits, ATR-scaled trailing, break-even, time-based exits, re-entry logic), full parameter exposure for Strategy Analyzer optimization, and data export pipelines for ML model training. MetaTrader 5 (MQL5): Expert Advisors for live and prop firm environments. Box breakout (6 versions), volatility compression straddle, Heiken Ashi exhaustion, multi-strategy EAs. Session filtering, spread guards, prop firm compliance (drawdown limits, daily loss caps, consistency rules, magic number management), and risk-based position sizing. Python ML Pipelines: XGBoost, LSTM, Transformer, and MLP models for session-type prediction, directional forecasting, and signal filtering. Walk-forward validation, SHAP explainability on every prediction, Monte Carlo stress testing (10,000 equity paths), and ONNX export for sub-millisecond inference inside NinjaTrader. Pine Script (TradingView): Multi-timeframe systems with non-repainting architecture, state-based logic, session-aware execution, and liquidity sweep / SFP detection. Cross-Platform: Strategy porting (NT8 ↔ MT5 ↔ ATAS ↔ DAS Trader ↔ Python), cross-platform signal routing across 5 simultaneous platforms, and broker API integration (Interactive Brokers TWS API, Rithmic, MetaTrader broker API). FLAGSHIP BUILD: IMBOne Cortex The most complex system in my portfolio — a quantitative trading intelligence platform I architected and built end-to-end as sole developer. WPF/.NET 9, 7 server projects, 30+ table MySQL database, SignalR real-time streaming, Redis caching layer. Core systems: real-time data engine processing 100K+ market data messages/min via Polygon WebSocket. Dark pool level detection (FINRA data, 4-tier classification). Gamma exposure engine (DIY from raw options chain data — zero vendor dependency). Pre-market scanner screening 8,000+ equities. ML pipeline with XGBoost classifiers (session-type F1=0.736, IB-direction F1=0.727), nightly automated retraining, SHAP explainability rendered in-app, and ONNX Runtime inference. 40+ engineered features. If I can build this solo, I can build what you need. SELECTED RESULTS ▸ Institutional Zone Breakout (NQ) — Sharpe 2.30, 53.9% WR, 668 trades ▸ ICT Liquidity Sweep — Sharpe 4.81, PF 1.93 ▸ Box Breakout EA (v6.5) — directional bias model, 82% AUC ▸ EURCHF ML Pipeline — walk-forward validated, mean Sharpe 4.02, all folds positive ▸ 80% Rule Debunk — 5+ years NQ data, measured 18.3% vs. claimed 80% ▸ Level Touch Probability — distance gravity model, AUC 0.78–0.85 ▸ NQ Breakout Pipeline — MBO-10 microstructure, absorption signal identification HOW I WORK Understand your strategy logic and goals before writing code Build clean, documented, configurable code — every parameter exposed Validate rigorously — walk-forward, Monte Carlo, statistical significance testing Deliver full source code + performance report + walkthrough session All client strategy code treated as confidential. NDA available. Your edge is yours. PROP FIRM EXPERIENCE: Systems designed for Tradeify, TopStepX, and similar environments. Trailing drawdown, daily loss limits, consistency rules. Risk controls enforced independently from strategy logic — the risk layer has absolute veto authority. TECH STACK: C# (27 yrs) · NinjaScript · MQL5 · Python · Pine Script · XGBoost · LSTM · SHAP · ONNX Runtime · SignalR · MySQL · Redis · Databento MBO-10 · Polygon API · Rithmic · IB TWS API · pandas · NumPy · scikit-learn PhD · 27 years C# · 17 years university teaching (I explain what I build — you'll understand your own system) · Bilingual English/Spanish · South Florida (EST) Let's build something that works in live markets, not just in backtesting.