5 days ago
AWS unveils Bedrock AgentCore & boosts generative AI with $100 million
Amazon Web Services (AWS) has announced new offerings to advance the use and management of AI agents, including the introduction of Amazon Bedrock AgentCore, new Marketplace solutions, and additional investment in generative AI development.
Amazon Bedrock AgentCore
At the centre of the update is Amazon Bedrock AgentCore, described as a comprehensive suite of services for building, deploying, and operating AI agents securely at scale. AWS states that the aim is to help organisations transition from AI agent experimentation to robust enterprise deployment, ensuring compliance with security, reliability, and governance standards.
The new AgentCore platform incorporates several components. AgentCore Runtime provides secure and dynamic support for interactive and asynchronous workloads, allowing agents to operate up to eight hours, which AWS claims is the longest in the industry. Session isolation and compatibility with any framework are also highlighted features.
AgentCore Memory supports both short-term and long-term memory functions for agents, aimed at enabling more context-aware responses with high accuracy. AgentCore Identity addresses authentication, allowing AI agents to securely access a range of tools and services while integrating with existing identity solutions like Amazon Cognito, Microsoft Entra ID, and Okta.
AgentCore Gateway gives AI agents a means to securely discover and utilise APIs, Lambda functions, and other resources, transforming them into agent-compatible tools. Meanwhile, AgentCore Code Interpreter enables secure execution of code within sandboxed environments, offering customisability around security and computational needs.
To facilitate agent interaction on the web, the AgentCore Browser Tool allows agents to navigate websites and complete forms at scale via a secure, model-agnostic cloud browser. For operational visibility, AgentCore Observability - powered by Amazon CloudWatch - provides dashboards and telemetry covering 81 metrics, supporting integration with established observability frameworks.
Early adopters, including Itaú Unibanco, Innovaccer, Boomi, Epsilon, and Box, are already utilising AgentCore in developing their AI agent solutions. "It's a tectonic change in a few dimensions," said Swami Sivasubramanian, AWS VP for Agentic AI. "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."
Marketplace expansion
AWS has launched a new AI Agents and Tools category in the AWS Marketplace. This addition is intended to simplify the process for organisations seeking commercial AI agent solutions, enabling customers to discover, buy, deploy, and manage offerings from various providers through a consolidated interface.
According to AWS, the Marketplace enhancements are designed to accelerate adoption by providing customers with streamlined access to professional services and pre-built tools required to build, maintain, and scale agentic systems.
Generative AI investment
AWS has pledged an additional $100 million to the AWS Generative AI Innovation Centre, bringing further resources to support the global push towards agentic AI systems. The centre's teams collaborate with customers and partners, focusing on productivity improvements and customer experience transformation through AI.
Several AWS customers have reported practical outcomes from working with these AI offerings. Warner Bros. Discovery Sports Europe has developed an AI-driven solution that assists cycling commentators in rapid fact-finding. The solution utilises Amazon Bedrock, Anthropic's Claude 3.5, and over 100 AWS services. BMW has also built an AI system on AWS to streamline diagnostics for over 23 million connected vehicles. Other companies cited include Syngenta and AstraZeneca, both reporting that agentic AI has contributed to changes in their operational models.
The global team supporting the Generative AI Innovation Centre comprises AI scientists, strategists, and engineers who work directly with customers to address technological challenges, especially those related to enterprise implementation and scale.
