FoodBot is an intelligent, LLM-powered chatbot that simplifies food discovery and ordering. It enables users to interact naturally, sharing their location and craving, while the bot fetches top nearby restaurants using SerpAPI, generates real-time menus via LLM tools, and maintains context through a Knowledge Graph. The architecture leverages LangChain’s ReAct agent paradigm, orchestrating multiple modular tools including a LocationNormalizer, RestaurantSearchTool, and MenuTool. It features a clean Gradio interface, blazing-fast Groq inference, and strong multi-agent design principles. This project showcases how conversational agents can solve real-world problems using fast, scalable, and modular AI systems. FoodBot also demonstrates how this agent-based design can be extended to other domains like travel booking and e-commerce marketplaces, making it a strong candidate for further development. details are in github repo: https://github.com/Ashwath-Shetty/raise_hackathon_prosus_track?tab=readme-ov-file#-how-to-extend-this-solution-to-other-services