In this session, we will explore the transformative power of Kubernetes and its self-healing properties in building a scalable, reliable, and cost-effective machine learning (ML) serving infrastructure on AWS. Drawing from my experience as the Director of Engineering at Onclusive, I will discuss the strategic decisions and practical steps taken to harness Kubernetes for optimizing ML deployment and operations. We will delve into key challenges faced, solutions implemented, and the resulting improvements in efficiency and cost savings. Attendees will gain actionable insights on leveraging Kubernetes to enhance their ML infrastructure, ensuring robust performance and scalability while minimizing operational costs.

 

Technical Level: Technical practitioner