AWS launches managed environment to deploy AI agents in minutes

Thursday 23 - 15:00

By: Dakir Madiha

AWS launches managed environment to deploy AI agents in minutes

Amazon Web Services has introduced a managed runtime environment designed to accelerate the deployment of artificial intelligence agents, reducing setup time from days to minutes. The new capability is part of Amazon Bedrock AgentCore, an infrastructure layer that removes the need for developers to build orchestration logic from scratch. The launch reflects growing demand for simplified tools to develop and scale AI-driven applications.

The managed system replaces complex workflows traditionally required to configure AI agents. Developers previously had to handle orchestration code, tool integration, memory configuration and compute management. With the new approach, deployment relies on a configuration driven model that requires only three API calls. Users define a model, a system prompt and a set of tools, while the platform manages reasoning, tool selection, execution and response streaming.

The runtime environment is powered by Strands Agents, an open source software development kit released by AWS in 2025. It is already used internally in services such as Amazon Q and AWS Glue. Each agent session runs in an isolated micro virtual machine with its own file system and shell access. The system also preserves session state, allowing agents to pause and resume tasks without losing progress. For teams requiring more control, the platform supports a transition from configuration based workflows to code driven orchestration without major architectural changes.

AWS has also introduced AgentCore CLI, a command line interface that consolidates the entire development lifecycle of AI agents into a single workflow. The tool supports project scaffolding, local testing, deployment and execution directly from the terminal. It is compatible with multiple frameworks, including Strands Agents, LangGraph, Google ADK and OpenAI Agents SDK, and works with models from Amazon Bedrock, Anthropic, Google and OpenAI.

In parallel, AWS is rolling out prebuilt AgentCore skills for coding assistants such as Amazon Kiro, Anthropic Claude Code, OpenAI Codex and Cursor. These integrations are designed to provide richer context within development environments, enabling more efficient coding workflows and improved agent performance.

The managed agent runtime is currently available in preview across four AWS regions, including US West, US East, Europe and Asia Pacific. The CLI is accessible in all commercial regions where AgentCore operates. AWS stated that there are no additional charges for using the runtime, CLI or prebuilt skills, with customers billed only for the underlying compute and service usage.

The release signals a shift toward infrastructure abstraction in AI development. By reducing technical barriers and simplifying deployment, AWS aims to enable faster experimentation and broader adoption of agent based systems across industries.