AI Image Search with OpenAI Embeddings & Vector Store

This n8n workflow builds a searchable image library by teaching a vector database to understand pictures through text. It pulls images from Google Drive, analyzes each one's color palette and content with a multimodal AI model to generate descriptive keywords, then combines that information into a document that's embedded with OpenAI Embeddings and stored in an in-memory vector store. The result is a practical way to find visually similar or thematically related images using plain-language search, without training a custom computer-vision model - perfect for photographers, marketers, and content teams managing large media libraries.

Category: RAG & Knowledge Base