What it is: Atlas Agent Engine is an enterprise-grade data, governance, and runtime engine for AI agents. It provides persistent memory, context, built-in guardrails, and state management so developers can move agents from prototype to production without getting bogged down in back-end data and infrastructure plumbing like custom state stores, guardrails, or data pipelines.

Who it's for: Atlas Agent Engine is for developers who want to run enterprise-grade AI agents in production, and focus on building agent logic unique to their business instead of spending time on underlying infrastructure.

Why it matters: Developers need total control to move agents seamlessly from prototype to production - with the flexibility to run anywhere, and the scalability that comes from a platform that already powers trillions of daily requests. And all without the hassle of stitching together custom state, memory stores, governance, or pipelines.

How to get started: Atlas Agent Engine is available in Public Preview on AWS. As a Public Preview, capabilities may change before General Availability. Learn more in our Technical Documentation.