GitLab acquires tool to use machine learning in code review

GitLab has acquired UnReview, giving it technology to leverage machine learning to find code reviewers. The acquisition is part of GitLab’s commitment to offer a broad DevOps platform.


GitLab is acquiring UnReview from and will integrate the tool into its DevOps platform by the end of this year , which will be available to the company’s software-as-a-service customers. UnReview is a tool that advises code reviewers to developers using machine learning.

The first version of the tool was released last year . UnReview taps into development platforms such as GitLab and GitHub to retrieve project data, such as review history, reviewer activity, and pull request content . After cleaning this data, a model trains algorithms so that the tool can provide suggestions on which reviewers are suitable for pull requests and what the workload will be, for example. After a match, UnReview visually displays the results, such as the interaction between reviewer and developer.

GitLab promises to extend the technology in the future with capabilities to automate other tasks, such as prioritization and problem assignment. GitLab’s chief technical officer, Eric Johnson, calls the acquisition “the first step in building GitLabs Applied Machine Learning for DevOps.” An acquisition amount has not been disclosed.