The terms MLOps, ModelOps, AIOps, and DataOps are used, to a great extent, interchangeably. It is important to know what MLOps is, and how it is different from ModelOps, AIOps, and DataOps before we move ahead with MLOps. In this chapter of MLOps, you will learn about the difference between MLOps, ModelOps, AIOps, and DataOps.
MLOps
MLOps refers to the practices and processes that are involved in the development and operationalization of Machine Learning models. The various steps that are involved are- developing, managing, deploying and integrating, and monitoring Machine Learning models.
ModelOps
ModelOps refers to the practices and processes that are involved in developing, managing, deploying and integrating, and monitoring Machine Learning models. It takes into consideration the development and operationalization of all types of models including Machine Learning models and other types of models as well, such as rule-based models and Knowledge Graphs. This makes ModelOps a super-set of MLOps, and MLOps a sub-set of ModelOps.
Since Machine Learning models, unlike other traditional software and other models are very dynamic, they call for a much different approach towards their development and operationalization.
AIOps
Artificial Intelligence for IT Operations or, AIOps refers to the use of Artificial Intelligence for enhancing IT Operations. AIOps leverages Machine Learning and Big Data to enhance IT Operations. For example, using Artificial Intelligence to predict the causes of an outage in the system and help in its resolution.
DataOps
DataOps refers to the practices and processes used to improve the quality of data while reducing the cycle time for Data Analytics. While the scope of DataOps is limited to the data, the scope of MLOps is much wider. However, both DataOps and MLOps share some common ground.
Although some of the practices of MLOps can be used for ModelOps. In this tutorial, our focus will be exclusively on MLOps.