Implementing ML pipeline: MLOps & GitOps principles

Machine learning operations (MLOps) aim to automate the entire lifecycle of the model, from development to deployment and management in production. Organizations can combine MLOps and GitOps concepts to achieve more efficient and controlled deployments and management of machine learning models.

In this article, Nataša Radaković and Nina Romanić, ProductDock’s software developers, together with Ivana Šenk, an associate professor at the Faculty of Technical Sciences, describe the machine learning pipeline that they created using available tools and following MLOps and GitOps practices. This paper was presented at the 19th International Scientific Conference of Industrial Systems – IS’23, held in October at the Faculty of Technical Sciences. Elevate your understanding of MLOps and GitOps principles in the original article.