Build a Generative AI Chatbot Using a Vector Database for Custom Data (RAG)

Nov 05, 20233,626 views14:39

In this video we'll build a generative AI chatbot that can answer questions based on your own custom data. We'll walk through how to use Retrieval Augmented Generation (RAG) to query OpenAI's LLM using a vector database, showing you how to set up the system and test the chatbot end-to-end.

Open in YouTube

Description

This video demonstrates how to use Retrieval Augmented Generation (RAG) to query OpenAI's Large Language Model (LLM) using a Vector database. This technique leverages the language understanding and summarization capabilities of Generative AI while introducing semantic understanding of our own data. In this video I walk through building a chatbot using OpenAI GPT models, Pinecone, and Python Notebooks. This video is a walk-through of a post available at https://robkerr.ai/generative-ai-chatbot-grounding-data-vector-text/ Quick Links: 0:00 Introduction 0:52 Setup 1:55 Source Data 3:14 Pinecone Index 6:43 Vector Embeddings 6:53 Encoding Content 9:54 Testing Vector Database 12:13 Testing Chatbot