Containers are a good fit for tasks that require flexibility in resources, while AI models demand significant computational capacity. And managing that capacity can be a daunting task. Sounds like a partnership in the making! The challenge is to avoid upfront investments in expensive hardware or reserved capacity and instead deploy AI models and the required infrastructure capacity on demand. The Kubernetes AI Toolchain Operator (Kaito) on Azure Kubernetes Service offers a solution. In this session, we'll discuss the provisioning requirements, and demonstrate how to effectively run AI models on Kubernetes.
Wesley is a Microsoft Azure MVP and is working as a Principal Azure Architect at Intercept. His day to day activities include designing, implementing and optimizing Azure solutions, focused on Cloud Native (Kubernetes), platform engineering, serverless and automating cloud operations. When he’s not working on a project, he’s giving a workshop, training or test-driving new Azure features. As a big advocate of Cloud Native, Infrastructure as Code and DevOps you will probably hear him talking about any of these subjects at least a few times a day (or hour). If you want to talk Azure, get in touch!