Graph retrieval augmented generation (Graph RAG) is gaining momentum as a powerful addition to traditional vector search methods. Graph RAG leverages the structured nature of graph databases, which organize data as nodes and relationships, to enhance the depth and contextuality of retrieved information. Knowledge graphs are great for representing heterogeneous and interconnected information in a structured manner, but constructing one can be challenging. To simplify this process, researchers have been experimenting with Large Language Models (LLMs), which can automate significant parts of the knowledge graph creation process. The code is available on GitHub, and it sets up a Neo4j instance to demonstrate how to construct a knowledge graph using LLMs. The Graph RAG approach combines structured graph data with vector search through unstructured text to achieve the best of both worlds. It uses a hybrid retrieval approach that combines keyword and vector indexes with graph retrieval for RAG applications, allowing users to pose questions and retrieve relevant information from the knowledge graph. The code provides examples of how to implement this approach using Neo4j and Python.