In this walkthrough, Zack from New Relic demonstrates how to eliminate the "Toggle Tax" by bringing live production telemetry directly into your AI agents. Usually, using AI assistants like Claude or Gemini for troubleshooting requires manual correlations between the code in your IDE and the performance data in your observability platform. This disjointed workflow slows down MTTR and forces engineers to bridge the context themselves. The New Relic AI MCP Server changes the game by using the Model Context Protocol to give your agents direct access to your telemetry. Now, your AI doesn't just guess code - it observes your system's reality. Key Takeaways: Zero-Distance Fixes: Watch Claude Code identify performance issues in a live trace and map the optimization directly to the local source code in one unified workflow. Democratized Insights: Learn how Atlassian Rovo allows first responders to query performance trends and service dependencies directly within Jira tickets - no NRQL expertise required. Unified Ecosystem: One open protocol supporting GitHub Copilot, Gemini Code Assist, and more, making New Relic the single source of truth for your entire AI stack. 0:00 The AI Assistant Blind Spot 0:17 Defining the "Toggle Tax" 0:24 Introducing the New Relic AI MCP Server 0:30 Beyond Chat: Deep-Linking into System Logic with Claude 0:50 Automated Root Cause Analysis & Code Optimization 1:01 Scaling Context: Atlassian Rovo & Democratized Insights 1:30 Summary: One Open Protocol, Every Agent 1:40 Eliminate the Toggle Tax & Get Started Today Eliminate the toggle tax and start solving today: https://newrelic.com/platform/product-preview --- Subscribe to the channel → https://www.youtube.com/c/NewRelicInc?sub_confirmation=1 Follow New Relic on: New Relic → https://newrelic.com Twitter → https://twitter.com/NewRelic LinkedIn → https://www.linkedin.com/company/new-relic-inc- Facebook → https://www.facebook.com/NewRelic

 202      5