AWS unveils AgentCore suite & invests USD $100 million in AI agents

AWS has introduced new tools and investments aimed at supporting organisations to build and deploy AI agents securely and at scale.

Central to these developments is Amazon Bedrock AgentCore, a suite of seven core services that enable companies to develop, deploy and manage enterprise-grade AI agents. The announcement also includes an expanded AWS Marketplace for AI tools and agents, and an investment of USD $100 million in the AWS Generative AI Innovation Centre to accelerate agentic AI development.

AgentCore features

Amazon Bedrock AgentCore provides a framework-agnostic set of services designed to support the transition from AI experimentation to enterprise deployment. The suite covers runtime, memory, identity, tool integration, code interpretation, web browsing, and observability, aiming to address key technical and governance needs for production-ready AI agents.

Swami Sivasubramanian, AWS Vice President for Agentic AI, outlined the company's strategy. He emphasised that AI agents - autonomous software systems that reason, plan, and adapt - could significantly accelerate productivity and outcomes in multiple industries. He said:

"It's a tectonic change in a few dimensions. It upends the way software is built. It also introduces a host of new challenges to deploying and operating it, and potentially most impactfully, it changes how software interacts with the world - and how we interact with software."

AgentCore's seven core services are designed to meet different operational aspects of running secure and scalable AI agents:

• AgentCore Runtime supports both low-latency interactive experiences and complex asynchronous workloads for up to eight hours, and provides session isolation in a framework-agnostic manner.
• AgentCore Memory offers long-term and short-term memory functionalities for agents to maintain context, aiming for high accuracy across memory types.
• AgentCore Identity integrates seamlessly with existing identity providers, such as Amazon Cognito, Microsoft Entra ID, and Okta, to secure and manage agent authentication.
• AgentCore Gateway enables secure access to a variety of digital tools, allowing easy transformation of APIs, Lambda functions, and existing services into agent-compatible resources.
• AgentCore Code Interpreter allows agents to execute code safely in sandboxed environments and can be tailored to meet varying security and instance requirements.
• AgentCore Browser Tool gives AI agents secure web access for automated online tasks, suitable for operations such as web form completion or navigating websites.
• AgentCore Observability leverages Amazon CloudWatch for real-time monitoring, dashboards, and telemetry, while integrating with customers' observability systems.

Customer organisations including Itaú Unibanco, Innovaccer, Boomi, Epsilon, and Box have begun utilising AgentCore as they seek to embed AI agents into their operations.

Secure & production-ready

According to Sivasubramanian, these services are aimed at helping customers as models and use cases evolve, allowing businesses to maintain secure deployment while scaling their AI capabilities. He highlighted the need for systems that can adapt as requirements change and said:

"Building systems that can act autonomously across digital boundaries, while maintaining the security, reliability, and governance standards required for enterprise deployment, is a critical challenge. AgentCore helps developers bridge the critical gap between proof of concept and production for AI agents. It delivers a set of composable solutions that allows organizations to move agents from prototypes to applications that can scale to millions of end-users."

Marketplace expansion

AWS is expanding its Marketplace to introduce a new category, AI Agents and Tools. This aims to help enterprises discover, purchase, and deploy AI agent solutions from a selection of providers. The intention is to streamline the process for organisations seeking to adopt AI agents, providing access to solutions and professional services for building, maintaining, and deploying agents at scale.

Sivasubramanian discussed this expansion during his remarks at the AWS Summit in New York, saying customers have access to a one-stop shop to accelerate the development of their AI initiatives. Customers leveraging these Marketplace solutions are expected to benefit from streamlined integration and delivery, with the support of specialist service providers.

Investment

AWS also announced a further USD $100 million for the AWS Generative AI Innovation Centre to support the development and deployment of agentic AI. This investment is designed to help businesses build and scale AI agents tailored to specific industry needs, supporting customers at various stages of their AI adoption journeys.

Additional enhancements to Amazon Nova were announced, offering customers improved customisation for building specific AI capabilities with higher accuracy and flexibility.

With these latest announcements, AWS aims to provide a more robust platform and ecosystem for organisations interested in deploying secure, scalable AI agents in production environments.