Enabling Natural Language AWS Operations with Amazon Bedrock AgentCore Runtime

Amazon Bedrock AgentCore Runtime has added support for the Model Context Protocol (MCP), enabling the construction of AI assistants that can operate AWS services using natural language. Traditional AWS management required navigating between the console, CLI, and multiple dashboards, but this new feature allows direct invocation of AWS APIs through a single conversational interface.

SREs and DevOps engineers previously had to check Amazon CloudWatch Logs, Amazon EC2 instance status, and IAM policies through separate interfaces during incident investigations. Capacity planning involved manually querying multiple services and combining the results.

(Reference: Integrating AWS API MCP Server with Amazon Quick using Amazon Bedrock AgentCore Runtime)

Workflow Automation through MCP Integration

With Amazon Bedrock AgentCore Runtime and MCP support, natural language queries can be directly converted into AWS API calls. Asking “Show me all running EC2 instances in us-east-1” yields accurate results immediately, without requiring tool switching or remembering API syntax.

Requests are executed securely within existing IAM permissions, with complete Amazon CloudWatch audit trails recorded for compliance. Instead of rebuilding connection logic for each workflow, AI agents can interact with AWS services through a single, reusable integration, standardizing their interaction method.

In Amazon Cognito authentication, an authorizer is set up during the CreateAgentRuntime operation, specifying the IdP-specific discovery URL and permitted clients. Existing agent code remains unchanged, and only the runtime deployment needs to be updated with the authorizer settings.

(Reference: Integrating AWS API MCP Server with Amazon Quick using Amazon Bedrock AgentCore Runtime)

Applying AI Agents to Medical Imaging Diagnostic Workflows

A radiology workflow optimization system using Amazon Bedrock AgentCore and Strands Agents SDK has been implemented. Traditional deterministic rule-based engines ignored critical context such as radiologist expertise, current workload, fatigue level, and case complexity.

Analysis of 2.2 million cases across 62 hospitals revealed that inefficient case assignments caused 17.7-minute delays for emergency cases, resulting in $2.1-4.2 million in costs across the hospital network. AI agents evaluate radiologist expertise, current workload, fatigue patterns, case complexity, clinical urgency, and availability simultaneously to assign cases optimally.

The system continuously learns from historical patterns and adapts to changing conditions, minimizing incentive structures that promote cherry-picking behavior. Radiology Partners recognizes this as a mission-critical workflow function and has partnered with AWS to adopt Agentic AI for intelligent workflow optimization.

(Reference: Intelligent radiology workflow optimization with AI agents)

Implementation Patterns for API Gateway Documentation

API Gateway REST API documentation allows adding and updating help content for individual API entities. Documentation can be performed using API Gateway REST API, AWS SDK, AWS CLI, or the API Gateway console. Importing and exporting documentation parts defined in external OpenAPI files is also possible.

When creating documentation parts, select the API in the Documentation type and use the following format in the properties map editor:

{
  "info": {
    "description": "Your first API Gateway API.",
    "contact": {
      "name": "John Doe", 
      "email": "john.doe@api.com"
    }
  }
}

The API Gateway console automatically stringifies JSON objects, so there is no need to encode the properties map as a JSON string. Documentation versions can be associated with API stages, and stage-specific documentation snapshots can be exported to external OpenAPI files.

(Reference: Document an API using the API Gateway console)

Summary

  • Amazon Bedrock AgentCore Runtime’s MCP integration enables natural language queries like “Show me all running EC2 instances in us-east-1” to be directly converted into AWS API calls, allowing infrastructure management without console switching.
  • Introducing AI agents into radiology workflows can solve traditional 17.7-minute delay and $2.1-4.2 million cost issues, building optimal case assignment systems considering expertise, workload, and fatigue.
  • Combining API Gateway documentation functionality with OpenAPI export enables automatic generation and distribution of comprehensive API documentation for developer portals.
  • Amazon Cognito authentication and IAM permissions enable secure automation of AWS operations by AI agents while maintaining existing security policies.