Using a search index to help AI understand queries

Weaviate Podcast describes an approach discussed in episode 133: extract entities from a query, check what exists in the search index and how it relates, then feed that context back before running the search. It frames this as retrieval-augmented generation (RAG) for query interpretation, followed by RAG on the results. The podcast also says the guests debate whether agentic search loops reduce the need for ever-better embedding models.