Travel AI Agent

This AI-powered travel assistant is a prototype that blends Retrieval Augmented Generation (RAG) and agentic LLM workflows to provide intelligent, real-time travel planning. Users can chat with the assistant to get destination recommendations, structured itineraries, and relevant travel data. The frontend, built with React, Three.js, and Web Speech API, offers a visually rich and voice-interactive experience. The backend uses LangChain, Pinecone, MongoDB Atlas, and Groq LLMs to manage semantic search, thread handling, and tool usage. This project showcases the practical application of generative AI and agentic systems in delivering user-centric travel solutions.

🔧 Tech Stack Highlights:

• 🖥️ Frontend: React 19, Vite, TailwindCSS, Three.js, and Web Speech API

• 🛠️ Backend: Node.js, Express, MongoDB Atlas, Pinecone, LangChain, LangGraph, Groq LLMs

• 🌟 Features: Conversational AI, semantic search, web scraping, multi-threaded chats, JWT authentication, 3D avatars