n8n Workflow: Multi-Format Document Processing for RAG Chatbot with Google Drive & Supabase

This n8n workflow automates the ingestion and processing of various document types from Google Drive into a Supabase Vector Store, making them ready for use in a Retrieval-Augmented Generation (RAG) chatbot. It handles different file formats, extracts their content, splits them into manageable chunks, generates embeddings using OpenAI, and stores these embeddings in Supabase for efficient retrieval.

What it does:

1. Monitors Google Drive: Triggers whenever a new file is added or an existing file is updated in a specified Google Drive folder.
2. Filters by File Type: Checks the file extension to determine if it's a supported document type (e.g., PDF, DOCX, TXT, CSV, JSON, HTML).
3. Downloads and Extracts Content:
   • For supported file types, it downloads the file from Google Drive.
   • It then extracts the text content from the binary file using a default data loader.
   • If the file is a CSV, it aggregates the data into a single string.
4. Splits Text into Chunks: Uses a Recursive Character Text Splitter to break down the extracted document content into smaller, overlapping chunks suitable for embedding.
5. Generates Embeddings: Creates vector embeddings for each text chunk using OpenAI's embedding model.
6. Stores in Supabase: Upserts the text chunks along with their generated embeddings into a Supabase Vector Store, making them queryable for a RAG chatbot.

Prerequisites/Requirements:

• n8n Instance: A running n8n instance.
• Google Drive Account: With credentials configured in n8n to access the target folder.
• OpenAI API Key: For generating text embeddings.
• Supabase Project:
  • A Supabase project with a configured pg_vector extension.
  • A table set up to store documents and their embeddings (e.g., documents table with content and embedding columns).
  • Supabase API URL and Service Role Key configured as credentials in n8n.

Setup/Usage:

1. Import the Workflow: Import the provided JSON into your n8n instance.
2. Configure Google Drive Trigger (Node 531):
   • Select your Google Drive credential.
   • Specify the "Folder ID" you want to monitor for new or updated documents.
   • Choose the "Operation" (e.g., "Watch for new or updated files").
3. Configure Supabase Vector Store (Node 1231):
   • Select your Supabase credential.
   • Enter the "Table Name" where documents and embeddings will be stored (e.g., documents).
   • Ensure the "Content Column" (e.g., content) and "Embedding Column" (e.g., embedding) match your Supabase table schema.
4. Configure OpenAI Embeddings (Node 1141):
   • Select your OpenAI credential.
   • Choose the desired "Model" for embeddings (e.g., text-embedding-ada-002).
5. Activate the Workflow: Once all credentials and configurations are set, activate the workflow.

Now, any new or updated documents in your specified Google Drive folder will be automatically processed and added to your Supabase Vector Store, ready to power your RAG chatbot!