Build self-service AWS Health analytics to find actionable health insights with AI agents powered by Amazon Bedrock

This article introduces Chaplin, an open-source solution that uses AI agents powered by Amazon Bedrock and the Model Context Protocol (MCP) to provide self-service AWS Health event analytics, enabling teams to ask questions in natural language and receive actionable insights without depending on AWS Support.

• Combines structured and unstructured data processing using multi-agent architecture to handle precise numerical analysis and contextual interpretation of health events
• Exposes AI-powered tools through MCP for integration with AI assistants like Claude Code and Kiro CLI, enabling conversational analytics directly in development environments
• Implements cost-optimized architecture using pattern-based classification for routine events and selective AI enhancement for complex analysis requiring contextual understanding
• Collects health events from multiple AWS accounts through centralized data pipeline with Amazon S3, DynamoDB, and Lambda for real-time processing and querying
• Provides three core capabilities: dynamic conversational analytics, multi-account data collection, and precise analytical processing combining exact counts with contextual insights
• Supports two deployment models: local install for individual developers and remote Lambda deployment for team-wide rollouts with automated infrastructure setup
• Enables proactive lifecycle management, cost optimization through early identification of deprecated services, and risk mitigation via security patch visibility

Chaplin transforms reactive AWS Health event management into proactive self-service analytics, reducing operational bottlenecks and enabling teams to independently plan migrations and assess business impact.