Portfolio

SELECTED
WORKS

Real projects, deployed and live. Each one built to solve a specific problem — not just to exist in a GitHub repo.

01 / 02
Personal AI Assistant
Python LangChain GEMINI 2.5 FLASH FASTAPI

Gem AI | Contextual API Engine

A high-performance, context-aware AI assistant designed for seamless full-stack integration. Built to function as a production-ready API, it moves beyond simple prompting by utilizing FastAPI and LangChain’s LCEL to manage sophisticated, session-based memory.

Core AI
LangChain (RunnableWithMessageHistory) orchestration paired with Gemini 2.5 Flash to maintain user identity and complex context across persistent sessions.
Deployment
Architected as a scalable backend engine on Render, featuring dedicated health-check synchronization to ensure zero-latency availability for client-side applications.
Live Demo ↗
02 / 02
MediMindAI Medical RAG System
Python LangChain Gemini Embeddings ChromaDB

MediMindAI | Medical RAG System

A specialized Medical Retrieval-Augmented Generation (RAG) system that provides grounded answers to health queries. Responses are retrieved directly from trusted medical datasets — dramatically reducing AI hallucinations and improving clinical accuracy.

Core AI
Gemini Embeddings for semantic search + ChromaDB vector storage for precision retrieval from medical corpora.
Architecture
Full RAG pipeline built with LangChain — grounded, hallucination-resistant, and domain-specialized.
View Project ↗

Next Project In Development

Working on a new agentic AI pipeline focused on advanced predictive analytics and autonomous decision-making. Check back soon or reach out to discuss.