This template automates the process of converting new Notion pages into embeddings and storing them in a Pinecone vector database. Whenever a new page is added in Notion, the workflow retrieves its content, filters out non-text elements, processes the text into chunks, generates embeddings using Google Gemini, and then inserts the vectors into Pinecone for semantic search and retrieval.
Perfect for building knowledge bases, AI assistants, and semantic search systems powered by your Notion content.
✨ Features
- Real-time Trigger: Automatically detects new pages in Notion.
- Content Extraction: Retrieves and concatenates page blocks into clean text.
- Filtering: Removes non-text content (images, videos).
- Text Splitting: Splits content into smaller chunks for better embedding efficiency.
- Metadata Creation: Stores useful metadata (page ID, title, created time).
- Embeddings with Google Gemini: Generates semantic vector representations.
- Vector Storage in Pinecone: Inserts processed embeddings into a Pinecone index for fast, scalable retrieval.
