Mohsin Ahmed Shaikh
AI/ML Support Team Lead, KAUST Supercomputing Lab
Abstract
MLOps is the practice of developing, deploying, and maintaining machine learning models reliably and repeatably. Kubeflow is an open-source platform, running on Kubernetes, that supports this practice from start to finish. In this half-day, hands-on tutorial, each participant takes their own model from first experiment to a running, monitored service, working on a realistic scientific use case in a ready-made cloud environment with GPUs. Through guided exercises, participants will learn to train a model with the best hyperparameter configuration, build pipelines that reproducibly retrain it on new data, and deploy it as an inference service. The tutorial suits beginners to MLOps, AI/ML developers, and researchers who want to move their models beyond the notebook. Basic Python is required; no prior experience with Kubeflow or Kubernetes is needed. Participants should bring a laptop with a modern browser, a GitHub account, and an email address for registration; all computing resources are provided.
Biography
KAUST) Mohsin Ahmed Shaikh is the AI/ML Support Team Lead at KAUST Supercomputing Lab (KSL), with over 16 years of experience in high-performance computing (HPC) and AI application support. He holds a PhD in Computational Bioengineering and completed postdoctoral research at the University of Canterbury, having previously served as a Supercomputing Applications Specialist at the Pawsey Supercomputing Centre. At KSL, Mohsin leads user enablement across the Shaheen supercomputer and Ibex GPU cluster, overseeing AI governance, agentic compliance workflows, and multi-tenant inference architectures. His expertise spans distributed deep learning, MLOps, containerization, and workflow orchestration using Kubernetes, Kubeflow, and Ray
Seats are limited
Tutorials run on Monday, 26 October 2026 at KAUST. Attendance is free, but registration is required.