Machine learning operations, or MLOps, is a set of practices to deploy and maintain machine learning models in production reliably and efficiently. This paper will explore the open source MLOps ecosystem as a whole and highlight tools of interest. MLOps tools will be compared and classified by task and maturity. Finally, there will be an overview of key players in the space. It seeks to highlight the knowledge necessary for practitioners and academics to select an open source machine learning operations tool (or tools) to suit their data science projects and machine learning systems.
The intended deliverable of this project is a paper that both lays out an overview of the open source MLOps ecosystem as a whole and highlights specific tools of interest. This paper evaluates and compares open source tools based on their functionality and maturity, and provides an overview of key players in the space.