Build and run AI workloads on Kubernetes using Kubeflow sdk #155
Description
Title of the talk/workshop
Build and run AI workloads on Kubernetes using Kubeflow
Abstract of the talk/workshop
Moving LLMs from a local notebook to a production-grade Kubernetes cluster introduces significant overhead: resource scheduling, GPU utilization, and pipeline orchestration. This session provides a practical blueprint for deploying AI workloads using Kubeflow sdk, thats a single python library to orchestrate all of this.
Attendees will leave with a clear understanding of how to run, fine tune and scale their models using Kubeflow components and ease of doing this using the single python sdk.
Category of the talk/workshop
Machine Learning, AI and MLOps
Duration (including Q&A)
30mins
Level of Audience
Intermediate
Speaker Bio
Akash Jaiswal, Software Engineer at Oracle
GSoC 2025 at Kubeflow and GSoC 2022 at CC Extractor
Speaker at 5+ events, regular contributor to Open Source
Work Ex: 1.6 years