@realcaseyrollins @curtis there is a practice called RAG (retrieval augmented generation) where they use a search system to find records and then feed those to the LLM to rewrite as an answer.

i don't trust it because it still relies on LLMs :blobcatglare2: