MongoDB integrates data platform with AI coding agents and automated embeddings

MongoDB, Inc. has announced new capabilities to integrate its intelligent data platform with AI development tools and improve retrieval accuracy for AI agents.

At the MongoDB Build Fest, the company introduced the MongoDB Atlas Managed MCP Server, which allows developers to connect AI agents to Atlas data natively. This integration supports tools such as Claude Code, Codex, Grok Build, and Devin, enabling agents to query, inspect, and update data without requiring additional infrastructure or manual connection strings.

Additionally, MongoDB is introducing automated embedding capabilities powered by Voyage AI models. This feature aims to simplify the process for developers by handling embedding and indexing automatically within MongoDB Atlas, eliminating the need for complex, separate embedding pipelines. The Financial Times is already utilizing these automated embeddings and Voyage AI models to improve the accuracy and cost-efficiency of its AI-powered semantic search across more than 100,000 daily searches.