Elastic, the Search AI Company, has introduced a new integration that brings comprehensive observability to AI agents built with Amazon Bedrock AgentCore. The integration enables developers and site reliability engineers (SREs) to monitor, optimize, and troubleshoot AI agents deployed on Amazon Bedrock AgentCore directly within the Elasticsearch platform. With this update, Elastic Observability captures all logs, metrics, and traces from AI agent workflows, providing a unified, end-to-end view of agent performance. The solution offers real-time insights into the full lifecycle of large language model (LLM) interactions, helping teams track agent behavior, control costs, ensure compliance, and maintain reliability across production deployments. “Organizations are rapidly moving AI agents into production, but observing every layer of the agentic AI stack remains a major challenge,” said Baha Azarmi, general manager of Elastic Observability. “Elastic’s Agentic AI Observability for Amazon Bedrock AgentCore delivers curated dashboards that map the entire LLM execution path, including AI guardrails for safety and compliance, and transparent cost tracking to help teams manage AI spend predictably.” The integration is designed to help engineering teams gain deeper visibility into how agents reason, make decisions, and interact with data. It supports proactive issue detection, performance tuning, and auditing—critical capabilities as AI agents become central to enterprise workflows. The Elastic Observability integration with AWS AgentCore is now available. Developers can begin using it immediately with a free 7-day trial on the AWS Marketplace. Elastic (NYSE: ESTC) is a global leader in search and AI-powered data solutions, enabling organizations to transform vast amounts of data into actionable insights. Its Search AI Platform serves thousands of companies worldwide, including over half of the Fortune 500. The platform powers search, observability, and security capabilities across industries. Elastic and its associated marks are trademarks or registered trademarks of Elasticsearch BV and its subsidiaries. All other company and product names are the property of their respective owners.