AI/ML Projects Portfolio
Cutting-edge applications in Artificial Intelligence, Machine Learning, and Data Science
M Capital: Multi-Agent Investment Platform
AI/LLMProject Overview
A multi-agent LLM system (LangGraph) in which 6 specialized agents research and debate to produce auditable investment recommendationa multi-agent LLM system (LangGraph) in which 6 specialized agents research and debate to produce auditable investment recommendations.
Key Features
- Convergence-gated debate plus an adversarial risk agent
- pgvector retrieval memory, Pydantic-enforced structured outputs
- Multi-provider LLM routing with token budgeting
- Real market-data integrations
RAG-based Tax Filing Chatbot
AI/LLMProject Overview
A sophisticated chatbot system that leverages Retrieval-Augmented Generation (RAG) to provide intelligent tax filing assistance. The system combines the power of GPT models with vector search capabilities for accurate and contextual responses.
Key Features
- RAG architecture for accurate information retrieval
- Real-time tax code updates and compliance checking
- NLP for complex tax queries
- User-friendly Streamlit interface
Proton Beam Energy Prediction
Deep LearningProject Overview
A DL model designed to predict proton beam energy for therapeutic applications in medical physics. This system helps optimize radiation therapy treatments by accurately forecasting beam characteristics.
Key Features
- Neural network architecture optimized for medical physics data
- Real-time beam energy prediction with high accuracy
- Flask-based REST API for easy integration
- Deployed on Google Cloud Platform for scalability
Cardiac Risk Identification
Machine LearningProject Overview
An XGBoost-based machine learning system for advanced cardiac risk assessment and prediction. This tool helps healthcare providers identify patients at risk of cardiovascular events through comprehensive data analysis.
Key Features
- XGBoost algorithm for high-performance risk prediction
- Feature importance analysis for medical interpretability
- Real-time risk scoring with confidence intervals
- Secure API deployment for clinical integration
Interested in Collaboration?
I'm always excited to work on innovative AI/ML projects that push the boundaries of what's possible.