At Proceedit Trading, I architected a 54-model ensemble across 16 strategies and developed a proprietary cumulative yield engine for backtest evaluation. The ensemble achieved 347% cumulative backtested yield on TSLA.
I led quant R&D and production pipelines for AI-based trading systems, and built live-like environments for model rotation, validation, and execution monitoring.
Earlier at Proceedit Trading, I designed machine learning and deep learning models for financial forecasting and market analysis. I also built custom genetic optimization algorithms and a master AI bot that orchestrated multiple trading bots through Flask APIs.
My projects include an NLP language detection system that achieved 0.98 accuracy, a CNN plant disease classifier that achieved 0.90 accuracy, and a genetic CNN-LSTM stock prediction model.
