Learn about K8sGPT / Codex MCP Integration and how it can support fault triage in K8s
Learn about Headlamp AI Assistant, an open-source AI assistant built directly into the Headlamp Kubernetes dashboard.
Change of program: live demo of K8sGPT and Codex MCP with Ron Petty, followed by an open roundtable on AI-assisted Kubernetes operations.
Maroon Ayoub (Senior Principal ML Engineer, Red Hat) presents llm-d, the Kubernetes-native distributed LLM inference framework and CNCF Sandbox project: KV-cache aware routing, disaggregated serving, and running large-scale inference on your own cluster.
Marcelo Yannuzzi (Principal Engineer, Outshift by Cisco) introduces CASA: intent-scoped authorization for Kubernetes multi-agent systems, enforced at the network layer.
The KubeRay maintainers walk through running distributed AI and ML workloads on Kubernetes with Ray: a quick intro to KubeRay for newcomers, then a deeper dive into a real-world use case and recent features.
Josh Halley (Cisco, and a CNCF AI TCG co-organizer) on confidential computing for AI agents and securing your runtime environment. Our first APAC-timezone session.
James Luan, CTO and cofounder of Zilliz and a Milvus maintainer, goes under the hood of vector search at scale: indexing internals, storage engine tradeoffs, and how Milvus fits into RAG and agent architectures on Kubernetes.
Kubernetes Storage SIG co-chair Xing Yang presents the CNCF TAG Infrastructure whitepaper on how storage architecture must evolve for AI workloads: covering training, inference, and AI agent storage patterns on Kubernetes.
Sympozium: A Kubernetes-Native Agentic Control PlanePart of the CNCF Artificial Intelligence Technical Community Group M...