Demo with Emamurho Ugherughe
summary
SAP runs machine learning models on Kubernetes while exposing them through Cloud Foundry as a service instance with service keys, using a custom cf CLI plugin that supports upload, deploy and describe. Training runs on NVIDIA GPUs, often as week‑long jobs on datasets in the gigabyte–terabyte range with data caching; trained artifacts land in a model repository. Inference containers—examples use TensorFlow Inception model—are deployed to Kubernetes and the demo shows successful image classification (dog) after the cf deploy reports succeeded.
AI-written summary · about
chapters (5)
- 0:00Context and platform choices
- 2:55Introduction of Emamurho Ugherughe
- 3:39SAP’s ML architecture overview
- 8:18Demo of ML service integration
- 11:48Q&A and wrap-up
related talks
46mDeploying scalable Machine Learning models with Kubernetes
42mDeploying ML at Scale with Kubernetes and TFX- 42mDeploying ML at Scale with Kubernetes and TFX
22mOperationalising Data Science on Cloud Foundry
40mStreet fighting techniques for multi tenant machine learning...
34mServing ML Models at Scale with Seldon and Kubeflow
31mKubeflow Intro – Michał Jastrzębski & Ala Raddaoui, Intel
54mEconomics and Best Practices of Running AI/ML Workloads on Kubernetes