What it is: Automated Embedding enables developers to build semantic search in minutes using Voyage AI models, without needing external pipelines or machine learning expertise. With a single click, vector embeddings are generated directly in the database and remain synchronized as the underlying data changes.

Who it's for: Atlas developers who want to build semantic search without managing external embedding pipelines or machine learning infrastructure. This is useful for teams building RAG applications, recommendation engines, or AI agents who need embeddings to stay in sync as data changes.

Why it matters: Building semantic search traditionally requires standing up separate embedding pipelines, managing model integrations, and keeping vectors in sync as data changes. Automated Embedding eliminates that overhead entirely. Embeddings are generated and updated automatically at the field level, in near real-time, with token usage and index health visible directly in the Atlas UI.

How to get started: Automated Embeddings is now available in MongoDB Vector Search on Atlas. Learn more in the documentation.