Tuesday, June 4, 2024 - 8:30am to 12:00pm

MLOps: DevOps for Machine Learning

Much attention is given to machine learning model training and testing in the industry. While these activities are essential for producing a production-ready machine learning model, organizations face some critical business challenges that must be addressed when building and testing machine learning models. Things like the reproducibility of results, accuracy of predictions, reusability of components, and trackability of experimentation are all vital to the success of any application.  The term MLOps has emerged as a method for applying DevOps practices and automation to the machine learning model development, testing, and delivery process. Join Jeffery Payne as he discusses what MLOps is all about and how to leverage DevOps practices and automation to solve the business challenges discussed above. Learn how to track your end-to-end machine learning development process so it can be reproduced as needed for governance and compliance. See what is necessary to keep models accurate in production and avoid bias and unfair predictions. Understand how to build reusable pipeline components to scale your machine-learning practices. Take home valuable ideas and tools for tracking your machine-learning experiments and overall process.

Jeff Payne

Jeffery Payne is CEO and founder of Coveros, Inc., a company that helps organizations accelerate software delivery using agile methods. Prior to founding Coveros, he was the co-founder of application security company Cigital, where he served as CEO for 16 years.

Jeffery is a recognized software expert and popular keynote speaker at both business and technology conferences on a variety of software quality, security, DevOps, and agile topics. He has testified in front of congress on issues such as digital rights mgmt., software quality, and software research.

Jeffery is the technical editor of the AgileConnection community (