DNS-like Trust Layer for Secure, Scalable AI-Agent Deployments on Kubernetes

Explore how to implement a DNS-inspired trust layer for secure AI agent deployments in this conference talk from MLOps World's GenAI Summit 2025. Learn about the Agent Name Service (ANS), an open protocol designed to address the critical gap in discovering, authenticating, and governing autonomous AI agents at enterprise scale on Kubernetes. Discover why traditional model-centric MLOps pipelines are evolving toward agent-centric ecosystems where AI agents autonomously retrain models, validate data, and remediate incidents with minimal human oversight. Examine the challenges of standardizing agent discovery and governance in production environments, and understand how ANS provides a solution through DNS-inspired architecture that enables trust, verification, and policy enforcement across the ML lifecycle. Watch live demonstrations showing ANS integration with GitOps workflows, Open Policy Agent (OPA) for policy enforcement, Sigstore for cryptographic verification, and open-source orchestration frameworks. Master practical implementation patterns for chaining validation, remediation, and notification agents while maintaining least-privilege access principles. Review performance benchmarks and security metrics from a production pilot system managing over 1,000 daily agent interactions at PayPal. Access practical templates, GitHub references, and architectural patterns for implementing ANS in your own Kubernetes-based MLOps infrastructure, complete with end-to-end trust verification from deployment through automated policy enforcement.