VedaAI
Built a full-stack AI-powered healthcare platform combining Ayurveda, generative AI, and machine learning for personalized health recommendations and disease risk prediction. Developed a LangGraph-based multi-agent architecture using RAG, OpenAI API, Gemini, and ChromaDB, and created REST APIs with Node.js, Express.js, and FastAPI using PostgreSQL and MongoDB.
TravAI
Designed an AI-powered travel planning platform using Node.js, React, and MongoDB for more than 50 users. Integrated Google Maps, Places, and Location APIs for real-time navigation and personalized recommendations, with backend services for itinerary generation and budget optimization.
InsightDoc AI
Implemented a multimodal RAG system using Python, LangChain, ChromaDB, Hugging Face Embeddings, and Groq Llama 3 for question answering over text documents and OCR-extracted images. Used hybrid semantic retrieval and entity-aware context filtering, and evaluated the pipeline with RAGAS metrics including 0.83 Context Precision, 1.00 Context Recall, and 1.00 Faithfulness.