Beyond the Green Dashboard

AI agents, hidden failures, and new attack surfaces are reshaping security

Recording Available Until December 29, 2026

Tage

Stunden

Minuten

Sekunden

🇬🇧 English

Rethinking Reliability and Security for Autonomous Systems

As AI agents become more autonomous, traditional DevOps, SRE, and security practices face new challenges. This live event explores polymorphic AI agents, silent system failures, and MCP security risks that conventional metrics and controls can overlook. Discover how engineering teams can adapt their practices to keep increasingly autonomous software reliable, observable, and secure.

From agents that can change their own behavior to systems that appear healthy while workflows quietly degrade, the sessions examine where established approaches reach their limits. Learn how workflow latency, cross system tracing, behavioral signals, and secure MCP deployments can help uncover hidden problems and address new attack surfaces.

Beyond Traditional Metrics: Building Resilient AI Systems

As software becomes more autonomous, reliability and security need to account for behavior, workflows, and interactions, not just infrastructure metrics. Explore how polymorphic agents can evolve beyond their initial instructions, how silent failures can disrupt workflows without triggering alerts, and how MCP servers introduce new considerations for supply chain and application security. Learn how observability, tracing, meaningful SLIs, and security controls can help teams detect hidden problems and build more resilient systems.

Your expert sessions

09:30 - 10:00 | When AI Agents Go Rogue: DevOps Lessons from the Rise of Polymorphic AI | John Willis

Explore how polymorphic AI agents can adapt their behavior, reconfigure themselves, and challenge assumptions about safe automation. This session uses real incidents and emerging research to highlight new risks and practical red flags for DevOps teams.

10:30 - 11:15 | Silent Failures: When Systems Are Up but Reliability Is Down | Abhimanyu Narwal & Mehdi Himmiche

Discover why traditional SLIs, SLOs, and golden signals can miss serious operational problems. Learn how workflow latency, cross system tracing, and behavioral patterns can reveal silent failures and strengthen reliability across both SRE and security environments.

11:45 - 12:30 | Agents and MCP Servers: Are the Electric Sheep Safe? | Brett Smith

Explore the emerging security risks of agentic AI and MCP servers in the software development lifecycle. Learn how MCP can support supply chain security, where new vulnerabilities emerge, and which mitigation strategies can help protect AI powered workflows.

Expert Knowledge for

DevOps and DevSecOps Engineers, who want to understand how autonomous AI agents change software delivery and security
SREs and Platform Engineers, who want to detect silent failures beyond traditional reliability metrics
Security Engineers, who want to identify and mitigate new attack surfaces created by AI agents and MCP servers
Software Architects, who want to build observable, resilient systems for increasingly autonomous environments

Join us and learn how to...

understand how polymorphic AI agents can change their behavior and introduce new risks
identify silent failures that traditional SLIs, SLOs, and golden signals can miss
use workflow latency, tracing, and behavioral signals to improve reliability
recognize security risks around agentic AI and MCP servers
develop practical strategies for more secure, observable, and resilient AI driven systems