Skip to content

Amazon Launches AgentCore for Building Multi-Agent Engineering Assistants

Amazon Web Services introduces Amazon Bedrock AgentCore, a platform for building, connecting, and optimizing multi-agent systems at scale.

Sources in the coveragePerformance and quality statements from a vendor are claims unless the coverage cites independent measurements.

Evidence & verification 3 statements

Specifications, source statements and editorial judgement have different scopes. A citation is not an independent performance test.

PlatformSource statement

Amazon Bedrock AgentCore

aws.amazon.com ↗Source observation date not recorded.
ArchitectureSource statement

Multi-agent graph

aws.amazon.com ↗Source observation date not recorded.
Knowledge BaseSource statement

Managed knowledge base

aws.amazon.com ↗Source observation date not recorded.

Amazon Web Services (AWS) announced Amazon Bedrock AgentCore, a platform for building multi-agent engineering assistants. This solution automates part substitution research by routing chat requests through a multi-agent graph, including a planner, specialists, and a supervisor, all hosted on AgentCore runtime.

Key facts

Fact Detail The source says
Platform Amazon Bedrock AgentCore “Amazon Bedrock AgentCore is a platform to build, connect, and optimize agents at scale, with any framework or model.”
Architecture Multi-agent graph “chat requests from the engineer’s application flow into a multi-agent graph hosted on AgentCore runtime”
Knowledge Base Managed knowledge base “specialists retrieve from a managed knowledge base that we created with Amazon Bedrock Managed Knowledge Bases”

What happened

Amazon Web Services (AWS) has introduced Amazon Bedrock AgentCore, a platform designed to build, connect, and optimize multi-agent systems at scale. This solution automates part substitution research by routing chat requests through a multi-agent graph, including a planner, specialists, and a supervisor, all hosted on AgentCore runtime. The specialists execute their tools and retrieve data from a managed knowledge base, while the supervisor synthesizes findings only when more than one specialist contributes. The system enforces human-in-the-loop rules, requiring an audit trail entry for every substitution decision. Amazon Nova Multimodal Embeddings, now available in AWS GovCloud (US), provides visual similarity scoring and chart retrieval capabilities, enhancing the assistant's functionality for aerospace and defense programs.


Source: Amazon Web Services

Keep following technology

Continue from here.

Explore more coverage or subscribe to Fathom Layer updates.

Back to the news desk ↗

Occasional verified technology updates. Unsubscribe anytime. See our privacy policy.