Smart Document Ingestion for Recursive Hybrid RAG with Google Drive & Supabase

This powerful workflow transforms raw documents into highly intelligent, context-aware data assets for your Retrieval Augmented Generation (RAG) systems. By continuously monitoring a designated Google Drive folder for new PDF or text files, it automatically downloads and processes content. The core innovation lies in its "Basic LLM Chain" which employs a Recursive RAG approach, using OpenAI's gpt-4o-mini to enrich each document chunk with broader context from the full document. This ensures that fragmented information, like incomplete numbers or table entries, is corrected and integrated, significantly enhancing the quality and relevance of embeddings stored in your Supabase vector database for superior AI search and response generation.