Restaurant Recommender Chatbot

Description

A RAG-based AI chatbot that helps users discover restaurants based on personal experiences. I log restaurants I visit into Notion with details like food, location, and a short summary. Using LangChain's Notion loader, the data is fetched and embedded with OpenAI embeddings, then stored in a vector database. When a user queries the chatbot, their message is embedded, a vector search retrieves relevant restaurant entries, and the chatbot provides contextually accurate recommendations. This project demonstrates practical experience with LLMs, Retrieval-Augmented Generation (RAG), vector databases, and building AI-powered chat interfaces.

Technologies Used

Key Features

  • Personalized recommendations based on user message
  • Retrieval-Augmented Generation (RAG) using vector search for relevant answers.
  • Vector database storing embeddings created with OpenAI for fast semantic search.
  • Notion Loader Integration: Fetches restaurant data from Notion for processing and embedding.
  • AI chat interface providing context-aware responses.