Episode 405 - RAG v/s Graph RAG with Prerna Singh

Retrieval augmented generation (RAG) systems are increasingly popular for developing document question-answering pipelines. This talk will begin by exploring the fundamental components of a RAG-based question-answering system. We will address the challenges these systems face and demonstrate how graph-based RAG systems can effectively address these issues. Additionally, we will examine state-of-the-art graph-based approaches and recent advancements in this field. The talk will provide a practical perspective on these approaches, particularly regarding computational cost.