Build & connect an incident investigation agent with Elastic Agent Builder and Gemini Enterprise

AI agents are most powerful when grounded in real organizational data. In this demo, discover how to use Elastic Agent Builder to create an incident-investigation agent powered by Elastic Observability data (logs, metrics, and traces) to quickly diagnose application performance regressions. You will also see how to expose the custom agent via the Agent-to-Agent (A2A) protocol and integrate it directly into Gemini Enterprise as a centralized platform. *What you'll learn in this video:* * Building specialized agents: How to use the GUI-based Elastic Agent Builder to define custom tools and runbook skills for automated troubleshooting. * Observability grounding: How the agent analyzes live logs, metrics, and traces to identify root causes (e.g., latency regressions in a frontend proxy) with full inspectability. * Agent-to-Agent (A2A) integration: How to export the agent card via an endpoint and register it within Gemini Enterprise. * Centralized Invocation: Querying and running Elastic-grounded workflows directly from the Gemini Enterprise web interface. Resources: Elastic Cloud on Google Cloud Marketplace → https://g.dev/cloud/marketplace Google Gemini Overview → https://g.dev/ai/learn-more-gemini Elastic Official Website → https://g.dev/ai/dynatrace-observability #Elastic #Gemini #GoogleCloud #DevOps #Observability #AIOps #CloudMonitoring #SRE 🔔 Subscribe to Google Cloud Tech → https://goo.gle/GoogleCloudTech Speaker: JP Hwang Products Mentioned: Elastic Observability, Elastic Agent Builder, Elastic Cloud, Google Cloud Gemini Enterprise, Google Cloud Run, Google Cloud Marketplace