MongoDB launches Atlas Agent Engine for production AI agents

MongoDB, Inc. has launched the Atlas Agent Engine, a unified execution, memory, and governance layer designed to facilitate the transition of AI agents from proof-of-concept to production environments. The engine aims to solve common engineering bottlenecks by providing a single platform for retrieval, persistent memory, and enterprise-grade security, reducing the need for teams to manually integrate disparate tools.

Key features of the launch include retrieval capabilities powered by MongoDB Voyage AI and a modular architecture that allows customers to adopt memory and governance layers independently. The development of this engine was supported by MongoDB’s expanding research and development hub in India, which has grown to approximately 110 engineers.

Additionally, MongoDB announced a partnership with Emergent, an AI software creation platform. This integration allows Emergent Pro and Enterprise customers to connect their existing MongoDB Atlas organizations directly to the platform, enabling them to build and run AI-generated applications using their own existing infrastructure and data models.