Agile MLOps: Designing AI for Continuous Improvement

The ML landscape is divers and evolving fast. As such, many attempts to formalize and clarify the principles and practices of MLOps fall short: irrelevant for some, or obsolete for others. These attempts focus on so called “best practices” and tools rather than measurable outcomes. In this talk we will review hard lessons learned working with hundreds of enterprise teams with ML production workloads and show how to design an MLOps project for continuous measurable improvement using the Agile principles.

Who is this presentation for?
Anyone who is putting AI/ML into production.

Prerequisite knowledge:
General understanding of ML development cycle.

What you'll learn?
How to design an MLOps project for continuous improvement using the Agile principles; how to prioritize, select the right tool and measure operational efficiency improvement.

Uri Rosenberg is the AI & ML Specialist Technical Manager for Europe, Middle East and Africa at Amazon Web Services (AWS). Based out of Israel, Uri works to empower enterprise customers to design, build and operate deep learning at scale. Before AWS, Uri led the ML projects at at&t innovation center in Israel, working on deep learning models with extreme security and privacy constraints. Uri is also an AWS certified Lead Machine Learning Subject Matter Expert and holds an MsC. in computer science from Tel-Aviv Academic college, where his research focused on large scale deep learning models.