I've built financial ML pipelines at Sarmad, producing 5- and 20-day ETF signals with walk-forward validation and deploying models through FastAPI and React. My work benchmarked CatBoost, LightGBM, XGBoost, LSTM, and Logistic Regression, reaching 0.861 ROC-AUC for neutral-versus-extreme classification.
I've also rebuilt social-media automation services at Andalusia Health and developed customer analytics models and a chatbot at ICTHub. Across hackathons and personal projects, I build LLM data agents, RAG learning platforms, computer-vision services, and multiplayer backend systems using Python, FastAPI, React, PostgreSQL, Docker, and cloud data tools.
