Hi guys, I'm trying to implement a RAG pipeline using DocumentSummaryIndex with Supabase + PGvector as vector DB, but unable to stote the doc store and vector store for retrieval. If anybody has implemented something similar, or if someone has used a vector store with DocumentSummaryIndex please help. Thanks!
Hi Logan, thank you for responding. Very new to this and apologies for sounding basic. Having difficulty thinking about how the doc store and vector store can be stored separately and then brought back together to create the document summary index object at the time of querying (and if that's even possible). Would greatly appreciate your thoughts.