In this tutorial, Matthew Berman introduces LangExtract, an open-source tool from Google designed to convert unstructured data into structured knowledge graphs. He explains how LangExtract utilizes large language models to perform entity and relationship extraction, making it particularly useful for creating Retrieval-Augmented Generation (RAG) systems. Matthew demonstrates practical examples of how to set up LangExtract, extract relevant data, and visualize the results, highlighting its applications in various fields, including healthcare and information retrieval.