I am creating a basic semantic store and retrieval Python application using LangChain. The storage part works fine and creates embeddings assigned to a new key in my documents. However, the retrieval part keeps returning to me:
Vector search found 0 semantic matches
I have created the search index using Quick Index in the Capella console and done the type mapping correctly, including assigning Vector type to the appropriate field. Exact SQL query matches are returning fine, but just no semantic matches are being returned. Played around with scoping and changing name of the parent field in my code, but to no avail.
Has anyone had similar issues and can give me any guidance? Thanks!