Overview

• EASI‑RAG outlines a method‑engineering framework with defined roles, activities, and techniques to guide SME teams in building RAG tools.
• In the reported validation, a RAG system answered operators’ questions using content extracted from operational procedures at a testing laboratory.
• The implementation was completed in under a month by a team without prior RAG experience and was iteratively refined based on user feedback.
• Authors say the approach enabled fast deployment, high user adoption, accurate answers, and improved reliability of the underlying data.
• The paper positions RAG as a way to curb LLM hallucinations and stale knowledge in resource‑constrained settings, while noting that broader validation and integration with fine‑tuned models are still needed.