Retrieval-augmented generation (RAG) systems coupled with knowledge graphs enhance accuracy in finding answers to queries.The article explores the use of LightRAG with Amazon Bedrock and Ollama to construct a knowledge graph from congressional records.Benefits include quick access to foundation models via Amazon Bedrock and offline embedding generation with Ollama.LightRAG integrates graph structures into text indexing, aiding in extracting information from legislative discussions effectively.The project utilized one day of congressional record data to build a knowledge graph for insightful queries.Instructions are provided to set up the environment, install dependencies, configure AWS credentials, and use the LightRAG CLI wrapper.LightRAG's source citing capability and customizability for Amazon's foundation models are highlighted in the article.Exploration of the generated knowledge graph visually is described as an engaging experience in the article.LightRAG is mentioned to still be under development, with improvements expected in entity and relationship outputs over time.Utilizing LightRAG and Amazon Bedrock for knowledge graphs improves data accessibility, cost-efficiency, and retrieval efficiency.