Data Science Projects Management, DataOps, MLOPs

This chapter discusses aspects related to Data Science project management that start with data collection and end up with the ML model deployment in a production environment. We discuss the importance of using research methods for effectively planning and managing Data Science projects. We also explain importance of applying best practices in DevOps for software development to Data Science projects that defined as DataOps, primarily focused on data management, and MLOps focused on ML based applications development. The chapter contains useful information about Data Science process models such as CRISP-DM and TDSP and explains how they can be used for Data Science projects management. Information about popular Machine Learning models such as PMML, ONNX and TensorFlow will help understanding the necessary steps in the transition from the model development stage to deployment and operation. The chapter finishes with a short overview of the Data Science and ML development platform by the major cloud and Big Data providers.