RAG: Retrieval Augmented Generation¶ It's the process of optimizing the output of a large language model.