Note: Ranjan Relan is an author and has published multiple courses (on Pluralsight and Coursera) on Azure including his course on MLOps (Optimizing Microsoft Azure AI solutions https://www.pluralsight.com/courses/microsoft-azure-optimizing-ai-solutions) which is a "partnership course with Microsoft" and is available as a linked reference in Microsoft Official Documentation.

About the Session
####################

In this session we shall discuss how data products which are powered by AI and machine learning algorithms can be industrialized using MLOps framework.

Challenges in industrialization of MLOps
*****************************************
Product-ionizing machine learning algorithms comes with its different sets of complexities as compared to rule based DevOps where the output is constant.

Guiding Principle of MLOps
**************************
We shall discuss what are the guiding principles of MLOps which includes how to ensure there is no change or minimal change in output, how to ensure there are no failures during model deployment, how to ensure data drift does not impacts the output,etc.

MLOPs & its Core Features
**************************************
We shall discuss what are the features of MLOps such as model versioning, data versioning, A/B testing, model deployment, continuous model monitoring, data drift,etc.

Enabling MLOps using Azure
#########################
Finally, we shall discuss how Azure services can be leveraged for enabling MLOps for data products and what all features are provided.