I'm a Full-Stack Developer passionate about leveraging AI/ML to solve real-world problems in healthcare. I specialize in building intelligent systems that bridge the gap between cutting-edge technology and practical clinical applications.
- 🔭 Currently working on AI-powered healthcare research agents and clinical decision support systems
- 🌱 Learning Agentic AI in Healthcare, Enterprise RAG Pipelines, HIPAA-compliant AI Design, and Healthcare AI Governance
- 💡 Interested in Healthcare AI, Autonomous Agents, Machine Learning, and Full-Stack Development
- 🎯 Goal: Build impactful technology that improves patient outcomes
- 📫 Reach me: [email protected]
AI-powered clinical decision support system using RAG + GPT-4
An intelligent healthcare platform that automates SOAP note generation, provides differential diagnoses, suggests ICD-10 codes, and recommends evidence-based treatment plans.
Tech: Python, FastAPI, OpenAI GPT-4, FAISS, React, RAG Architecture
Key Features:
- 🧠 RAG-powered semantic search through clinical knowledge base
- 📋 Automated SOAP note generation
- 🔍 AI-driven differential diagnosis with risk stratification
- 💊 Automatic ICD-10 medical coding
⚠️ Patient safety checks (allergy contraindications)
RAG-powered healthcare research assistant built at Elastic x Contextual AI Hack Night
A healthcare AI agent that searches authoritative medical literature (NIH, CDC, WHO) and provides evidence-based answers with source citations. Unlike regular LLMs that rely on stale training data, MedAssist uses up-to-date documents and cites every claim back to the exact source and page.
Tech: Python, Contextual AI Platform, Claude Sonnet 4.5, RAG Pipeline, Agent Composer
Key Features:
- 📄 RAG pipeline grounded in real healthcare documents — no hallucination
- 🔍 Semantic search + reranking across clinical guidelines
- 🔄 Multi-pass agentic research loop (up to 5 search rounds per query)
- 💬 Multi-turn conversations with follow-up support
- 📋 Automatic source citations with document name and page number
- ⚕️ Built-in medical disclaimers for responsible AI
Comprehensive AI agent security with 8 defensive layers
Multi-layered security system for AI agents including device fingerprinting, behavioral monitoring, prompt injection detection, memory integrity, and audit logging.
Tech: Python, Security Architecture, Prompt Injection Defense
current_focus = {
"Agentic AI in Healthcare": [
"AGENT Framework - human-first vs. agent-first workflow evaluation",
"HIPAA-compliant AI solution design",
"Healthcare AI governance policies",
"Strategic healthcare AI deployment"
],
"AI/ML": [
"AI Agents", "Advanced RAG Architectures",
"Enterprise RAG Platforms (Contextual AI)",
"Vector Databases", "LLM Fine-tuning",
"Semantic Search & Reranking"
],
"Agent Frameworks": [
"LangChain", "AutoGen", "Claude Agent SDK",
"CrewAI", "Contextual AI Agent Composer"
],
"Backend": ["Microservices Architecture", "GraphQL", "Real-time Systems"],
"DevOps": ["Kubernetes", "CI/CD Pipelines", "Cloud Deployment"],
"Healthcare": [
"FHIR Standards", "HL7", "HIPAA Compliance",
"Clinical Decision Support Systems",
"Healthcare Data Interoperability"
]
}📚 Upcoming: Harvard Data Science Review - Agentic AI in Healthcare Intensive (Feb 2026)
- 🏥 Healthcare AI Development: Building RAG-powered clinical decision support systems and healthcare research agents
- 🤖 AI Agents & Automation: Developing intelligent autonomous agents with LangChain, Claude, and Contextual AI
- 🔧 Full-Stack Engineering: End-to-end application development with modern tech stacks
- 🧠 LLM Integration: Implementing GPT-4, Claude, embeddings, rerankers, and vector search in production
- 📊 Data Engineering: Working with clinical data, medical coding, and patient analytics
- 🔐 Healthcare Compliance: Designing AI solutions with HIPAA compliance built in from the start
