If you’re trying to improve the quality of your LLM-generated responses, you’ve probably explored retrieval augmented generation (RAG).