Talk To Your Dashboards - Using MCP and LLMs To Simplify Observability

Learn how to simplify observability dashboard management in large-scale application services using Model Context Protocol (MCP) and Large Language Models (LLMs) through natural language interfaces. Discover the challenges of maintaining complex dashboards as systems grow, including outdated configurations and query drift over time. Explore how LLMs can streamline dashboard and alert management using natural language prompts, and understand the scaling challenges of integrating this capability into open-source systems due to fragmented APIs. Master the Model Context Protocol (MCP) as a solution that exposes observability primitives—dashboards, datasources, and alerts—through a standardized schema that LLMs can comprehend. Watch practical demonstrations of how MCP servers and clients can bridge LLMs with dashboard and alert management tools via natural language commands. Gain hands-on techniques for integrating LLMs into your observability visualization stack, specifically with tools like Perses, without requiring custom wrappers or brittle integrations. Leave with actionable strategies to implement natural language interfaces for your observability infrastructure, making complex dashboard management more accessible and efficient for your team.