Explore a comprehensive live coding session from NDC London where Phil Nash demonstrates how to build a Retrieval Augmented Generation (RAG) system from scratch. Learn the fundamental components that make RAG systems effective at providing large language models with up-to-date information while reducing hallucinations. Through hands-on coding, discover the inner workings of vectorization, similarity search, embedding models, and vector databases. Follow along as these components are assembled into a functioning augmented chatbot. This 53-minute talk provides valuable insights into the architecture behind successful chatbots and explains why these systems work, offering a solid foundation for developers interested in AI-powered conversational interfaces.