Cloud Native + Kubernetes AI Day Europe 2025
2025-04-01
London, United Kingdom
as the organisers put it
Cloud Native & Kubernetes AI Day Europe has officially wrapped! Thank you to all the attendees who joined us. Watch keynotes and all breakout sessions on the CNCF YouTube Channel.
who it's for
platform engineers, machine learning engineers, site reliability engineers, kubernetes operators, MLOps engineers, data engineers, AI platform architects
topics
large language model inference; retrieval augmented generation; vector databases (Milvus/Postgres); LLM instance gateways; KServe local model cache CRD; topology-aware scheduling (Kueue); Ray distributed inference with VLLM; WasmEdge edge inference runtime; GPU dynamic allocation (MIG/DRA); cluster GPU health monitoring; prompt injection and jailbreak mitigation
| title | author | length |
|---|---|---|
| Welcome + Opening Remarks | Thiago Bittencourt Gil, Yuan Tang | 06m |
Opening remarks set the conference frame that running AI on Kubernetes is normal now and the urgent problem is efficiency not feasibility, tracing evolution… | ||
| The State of GenAI & ML in the Cloud Native Ecosystem | Alejandro Saucedo, Bartosz Ocytko | 23m |
Overview of production GenAI and ML at scale with concrete claims: enterprises like Zalando run forecasting, pricing, search and a conversational fashion… | ||
| Lightning Talk: Scalable and Observable RAG for Question-Answering in a Box | Selvi Kadirvel | 13m |
Builds a self-hosted RAG stack on Kubernetes using Elotl’s Luna and Nova for autoscaling, VLLM on Ray for distributed inference, and Microsoft 53 plus Rubra AI… | ||
| ⚡ Lightning Talk: Introducing LLM Instance Gateways for Efficient Inference Serving | Abdel Sghiouar, Daneyon Hansen | 16m |
A Gateway API extension called LLM Instance Gateways exposes model-aware routing by extracting the model name from request bodies and steering traffic via an… | ||
| Serving the Future: KServe’s Next Chapter Hosting LLMs & GenAI Models (with Fun Drawings!) | Alexa Nicole Griffith, Tessa Pham | 29m |
KServe now ships concrete features for GenAI: a local model cache CRD that auto-downloads models to node PVs and self-heals, OpenAI protocol endpoints for… | ||
| Manage Cloud Native LLM Workloads Across Edge and Cloud Seamlessly Using KubeEdge and WasmEdge | Vivian Hu, Fei Xu | 25m |
Edge nodes run tiny models locally and automatically escalate to cloud models when accuracy drops, using WasmEdge runtime (<50MB) for portable, dependency‑free… | ||
| Engaging the Kubeflow Community: Building an Enterprise-Ready AI/ML Platform | Yuan Tang, Andrey Velichkevich, Andreea Munteanu, Johnu George, Ronen Dar | 44m |
Kubeflow now stitches operators and tools like Knative, Istio, KServe, the Trainer operator, KFP pipelines, model-registry and Spark operator into a composable… | ||
| Accelerate Your AI/ML Workloads With Topology-Aware Scheduling in Kueue | Michał Woźniak, Yuki Iwai | 24m |
Kueue adds topology-aware scheduling by consuming a simple node-label topology API linked to resource flavors, reconstructing a topology tree, running two… | ||
| Cluster Management for Large Scale AI and GPUs: Challenges and Opportunities | Claudia Misale, David Grove | 24m |
Runs Kubernetes-native tooling—Prometheus, Grafana, DCGM exporter, GPU Operator, ML Batch/Q and AppWrapper—across multi-node H100 GPU clusters with NVMe and… | ||
| Objection! AI Security Mistakes on Trial With Kubeflow and Confidential Computing | Annie Talvasto, Karl Ots | 26m |
Four recurring failures surface with concrete fixes and demos: untrusted model artifacts like pickle remote code execution that demand provenance checks and… | ||
| From Toil To Triumph: Harnessing Agentic AI To Streamline Infrastructure as Code | Jodee Varney | 32m |
Builds a GitHub-integrated reviewer for OpenTofu/Terraform pull requests that runs on a production copy of a repo and evaluates across 700+ pull requests… | ||
| AI, CERN, and the Quest for GPU Custody: How CERN Leverages DRA for Efficient GPU Sharing | Diana Gaponcic, Jan-Philip Gehrcke | 31m |
CERN centralizes GPUs into a shared pool and reduces idle time by combining time-slicing, NVIDIA MPS, and MIG (up to seven isolated instances) with Dynamic… | ||
| Open Source Tools To Empower Ethical and Robust AI Systems | Vicente Herrera, Alberto Rodríguez Fernandez | 26m |
Shows a hands‑on survey of open source tools with examples and artifacts: Garak and Gizkar for automated safety probes, LLM Guard and LLM Canary for runtime… | ||
| Lightning Talk: Benchmarking Your Distributed ML Training on the K8s Platform | Liang Yan | 12m |
A Kubernetes-native benchmark for distributed ML training that automates environment control (PVC, daemonset kernel/config tweaks), launches MPI jobs and… | ||
| Lightning Talk: {”spec”: “nodeSelector”: "EuroHPC"} | Diego Ciangottini | 12m |
Claim: a Kubernetes pod with nodeSelector: "EuroHPC" and SLURM/MPI settings in pod annotations runs as a job on EuroHPC supercomputers. Implementation uses… | ||
| Closing Remarks | Rajas Kakodkar, Ricardo Rocha | 01m |
Event closes for Cloud Native + Kubernetes AI Day Europe 2025 with thanks to Mirantis platinum sponsor and the AV team, panels, and attendees; organisers note… | ||
| nothing matches | ||