Automating RAG Indexing from Google Drive with n8n

This workflow automates ingesting Google Drive files (Google Docs and XLSX) into a Supabase/Postgres-backed RAG pipeline. It detects file updates in a monitored folder, downloads and extracts text or table rows, summarizes and splits content into chunks, and generates OpenAI embeddings (text-embedding-3-small) via Lang Chain nodes. The flow upserts document metadata and table rows into Postgres/Supabase, inserts embeddings into a Supabase vector store for semantic search, and removes previous entries to keep the index in sync.