Designing AI Workflows in the Cloud with AWS

Cloud services have become a key foundation for modern AI applications in industry. From scalable compute and storage to serverless workflows and managed services, the cloud enables teams to build, experiment with, and deploy AI solutions faster and more flexibly.

Unlike the research environment most students are familiar with, where work is typically done on local machines or high-performance clusters using tools such as Linux and SLURM, the cloud introduces a different paradigm. Instead of managing static infrastructure, developers can access scalable, on-demand resources and integrate managed services, fundamentally changing how AI workflows are developed, deployed, and automated in practice.

In this workshop, we will explore the fundamentals of cloud computing, discuss why cloud services have become essential for modern AI systems, and examine how different categories of AWS services fit together. We will cover services such as EC2, Lambda, S3, and other key building blocks commonly found in real-world cloud architectures.

Following a short introduction to core cloud concepts, participants will work in small groups on a hands-on architecture challenge. Each group will receive an AI-related use case and a set of AWS service icons. The goal is to design an appropriate cloud workflow, connect the relevant services, and present the reasoning behind the architectural decisions.