From Noise to Signal: Using MCP Servers for AI-Driven Alerting and Monitoring - Devyani Kota

This talk explores the limitations of traditional monitoring and alerting systems, which often generate excessive noise while lacking essential context. The speaker presents the Model Context Protocol (MCP), a framework that enhances observability workflows by incorporating structure, context, and reasoning. Attendees will witness how MCP can transform alerting and incident management into proactive systems through real-world examples and live demonstrations. The session highlights how contextual data from alert metadata, recent deployments, logs, and runbooks can be utilized by AI models to produce responsive agents capable of explaining incidents effectively. Participants will also learn to integrate AI agents into observability pipelines and see an AI-powered Alert Assistant in action, along with practical use cases and templates for production adaptation.