Kubernetes AI Summit with Kubeflow

AI is advancing at an unprecedented pace, from Large Language Models (LLMs) and Graph RAGs to new frameworks for ethical and responsible innovation. At the center of this evolution is Kubeflow, the cloud-native ecosystem enabling data scientists, ML engineers, and AI practitioners to build, deploy, and scale AI/ML workloads.

This event brings together the Kubeflow community, MLOps professionals, and AI enthusiasts to explore real-world patterns, cutting-edge tools, and production-ready solutions. Whether you’re new to AI or an experienced practitioner, you’ll gain valuable insights, hands-on knowledge, and connections that will shape the future of AI in the cloud-native world.

This event is one of our KubeCon + CloudNativeCon Europe CNCF-hosted Co-located Events. In-person attendees have the option to register for an All-Access In-Person KubeCon + CloudNativeCon pass that will include entry to ALL CNCF-hosted co-located events + KubeCon + CloudNativeCon.

Please check back — registration will open in October.

Contact sponsor@cncf.io to secure your sponsorship today! Signed contracts must be received by Monday, 2 February 2026!

Submit a proposal to speak! Submissions are being accepted through Sunday, 2 November 2025.

Click above to submit a proposal to speak at Kubernetes AI Summit with Kubeflow, or one of the other CNCF-hosted Co-located events.

Kubernetes AI Summit with Kubeflow topics include:

• LLMs & Generative AI
• Kubeflow Ecosystem & Community
• RAG & AI Agents
• End-to-End AI/ML in Production (MLOps/AIOps)
• Edge AI & Specialized Environments
• Ethics, Compliance & Security
• User Success Stories & Use Cases

Submission Types:

• Presentation: 25 minutes with 1 or 2 speakers presenting on a topic
• Panel Discussion: 35 minutes of discussion amongst 3 to 5 speakers (In an effort to promote speaker diversity, CNCF does not accept submissions with all-male panels, and three different organizations must be represented).
• Lightning Talk: A brief 10-minute presentation with a maximum of 1 speaker