"we devised a system leveraging Retrieval Augmented Generation (RAG) to greatly improve the quality of the structured output that represents such workflows." is not the same as reducing hallucination within the LLM. Please re-read my article again ;)