Fine-tuned Llama 3.1 8B · QLoRA · RAG serving stack · FastAPI · Pinecone · Azure

This project is a medical chatbot leveraging AI and NLP to assist users with medical-related queries. It utilizes Pinecone for vector search, FastAPI/Flask for backend, and a clean HTML/CSS frontend for an interactive UI. The chatbot is designed to provide quick and reliable responses using machine learning models and advanced search techniques.

This project is a Retrieval-Augmented Generation (RAG) app that pulls relevant Reddit posts about Elden Ring builds. It aims to generate helpful answers that can help players build upon the evolving meta of the game. The idea is to provide insights and strategies that are constantly updated based on community discussions.

A collection of GenAI experiments, implementations, and trial-and-error explorations. This repository documents various models, techniques, and applications, including bots, assistants, and other AI-driven projects. Whether you're looking for inspiration, debugging insights, or full-fledged implementations, you'll find a mix of structured projects.

An AI-powered Second Brain that lets you chat with your Obsidian notes. Built with Python, LangChain, ChromaDB, and Google Gemini 2.5.

Generic MCP server for Pinecone Assistant API with AI-powered RAG, strategic multi-search, YAML-configurable domains, token optimization, and cross-MCP integration. Reference implementation: USPTO patent examination (MPEP)