Automate Notion Content Ingestion to Pinecone Vector Store

Overview

This n8n workflow automates the process of transferring content from Notion to a Pinecone vector store. It's designed to capture new pages added to a specified Notion database, clean the content by filtering out non-textual elements, and then process this text for vectorization. The workflow leverages the Notion - Page Added Trigger to initiate the process, Notion - Retrieve Page Content to fetch the page's text, a Filter Non-Text Content node to ensure only relevant text is processed, Summarize - Concatenate Notion's blocks content to combine block text, and a Token Splitter to break down large texts into manageable chunks. Finally, Embeddings Google Gemini generates vector embeddings, and Pinecone Vector Store stores these embeddings along with metadata from the Notion page, making your Notion knowledge base searchable via vector similarity.