Kubernetes AI Day Europe 2022
as the organisers put it
Join us for Kubernetes AI Day to delve deeper into deploying AI at scale using Kubernetes.
who it's for
mlops engineers, platform engineers, kubernetes operators, machine learning engineers, data engineers, site reliability engineers, research scientists
topics
KServe model serving; Kubeflow Pipelines authoring; Argo Workflows with Katib; ModelMesh model caching; Ludwig AutoML with Ray; edge model debugging telemetry; fractional GPU and elastic training; multi-cloud Spark on Kubernetes; component registry for pipelines; transfer learning for vision
| title | author | length |
|---|---|---|
| Opening Remarks | Thiago Bittencourt Gil | 05m |
Brazil handles more than 45% of global credit card fraud, and the talk argues that without Kubernetes for scalable countermeasures large-scale fraud detection… | ||
| Challenges and Opportunities in Making AI Easy and Efficient with... | Maulin Patel | 10m |
Kubernetes (noted as Kuberenetes in the event listing) is framed as a strong platform for ML/HPC because it brings portability from laptop to cloud… | ||
| Kubernetes + AI Joining Forces in the Battle Against Cancer | Wojciech Małota-Wójcik | 27m |
Uses stardust to segment nuclei, extracts 11x11 patches, compresses each to a 10‑dim latent with an autoencoder (training error 4.7 per pixel ≈ 12 gray… | ||
| Efficient AutoML with Ludwig, Ray, and Nodeless Kubernetes | Anne Marie Holler, Travis Addair | 28m |
Ludwig AutoML uses Ray Tune and heuristics derived from 12 training datasets to pick architectures (TabNet for tabular) and narrow hyperparameter ranges… | ||
| Exploring ML Model Serving with KServe (with fun drawings) | Alexa Nicole Griffith | 28m |
KServe is a Kubernetes CRD that actually deploys and monitors inference services using Knative/Envoy (Istio) for routing, supports scale-to-zero with GPU/CPU… | ||
| Building AutoML Pipelines With Argo Workflows and Katib | Andrey Velichkevich, Johnu George | 22m |
Shows a concrete Katib+Argo setup that runs each Katib trial as an Argo Workflow, with the trial template carrying a full Argo YAML, the metrics collector… | ||
| Computer Vision Dog Breed Classification with Convolutional Neural Networks, TensorFlow and Kubeflow | Konstantinos Andriopoulos, Dorothea Kalliora | 29m |
A Jupyter notebook pipeline uses TensorFlow, transfer learning with ResNet-50 82% and VGG16 72%, plus a small CNN trained from scratch (much lower accuracy)… | ||
| Debugging Machine Learning on the Edge with MLExray | Michelle Nquyen | 27m |
An end-to-end debugging platform that instruments deployed edge models via a small API and writes per-inference logs of per-layer latency and tensors… | ||
| Enhancing the Performance Testing Process for gRPC Model Inferencing at Scale | Ted Chang, Paul Van Eck | 32m |
Shows a Kubernetes setup using KServe + ModelMesh to host high-density gRPC inference, with ModelMesh paging models in and out of memory as an LRU cache and… | ||
| A Component Registry for Kubeflow Pipelines | Christian Kadner | 29m |
Proposal for a centralized component registry API that stores Kubeflow Pipelines components as unified YAML packages with versioning and tags (Docker-like)… | ||
| Managing Multi-Cloud Apache Spark on Kubernetes | Ilan Filonenko, Aki Sukegawa | 31m |
Bloomberg runs Apache Spark across bare metal, EKS and AKS by moving executor creation out of user pods into an admin controller that reconciles a SparkApp… | ||
| Closing Remarks | Alex Collins, Jessica Andersson | 05m |
Thanks and acknowledgements for presenters, paper authors, the program committee Kristen Thiago Thiago Bob Luciano, and local AV crew Robert and Roger, plus… | ||
| A Deep Dive into Kubeflow Pipelines | Senthil Raja Chermapandian | 32m |
Overview of Kubeflow Pipelines architecture with concrete details: the pipeline service parses Python DSL and compiles YAML, metadata sits in a MySQL metadata… | ||
| nothing matches | ||