Interesting how RAG has become synonymous with Q&A chatbots, when it's really a pattern for any case where external context would help an LLM make better decisions. We're barely scratching the surface of applications of LLMs in engineering.