Hey everyone,
I was wondering if guys had any tips on how to synchronize your main database and the pinecone database / vector database… For my part, I have a backend in golang, and every time a text is inserted/updated/deleted in a table, I call an after insert/update/delete hook that deal with the embeddings and do the same thing in the vector database, and keep the same id. is it better to not duplicate, and keep the text in pinecone as a metadata (not in the main DB), or just stock the ids and vectors in pinecone that matches with the ones in the main DB ?
for the real geeks of golang I am using ent 