GitLab UnReview is open DevOps + Machine Learning (DevMLOpenOps)

A lot goes on at GitLab.

Among the work executed over the summer season was the acquisition of UnReview, a machine learning (ML) based technology for automatically identifying appropriate expert code reviewers and controlling review workloads and distribution of knowledge.

This acquisition is expected to advance the user experience within GitLab’s Dev Section including Manage, Plan and Create stages by improving a user’s ability to perform impactful code reviews by using ML to recommend code reviewers based on their previous contributions to areas of code as well as current reviewer workload. 

With this technology on board, GitLabs is now promising the capability for teams to be able to increase their velocity, code quality and security.

DevOps, pls meet M

We know DevOps is aligning to a whole lot of ML, but how much?

The GitLab 2021 DevSecOps survey suggest that some 75% of respondents report their DevOps teams are either using or planning to use ML/AI for testing and code review. 

Additionally, a majority (55%) of operations teams report their life cycles were either completely or mostly automated.

 “Integrating UnReview’s technology into the GitLab platform marks our first step in building GitLab’s Applied Machine Learning for DevOps,” said Eric Johnson, CTO of GitLab. “By continuing to incorporate machine learning into GitLab’s open DevOps platform, we are improving the user experience by automating workflows and compressing cycle times across all stages of the DevSecOps lifecycle. We’re also building new MLOps features to empower data scientists.” 

Image source: GitLab DevSecOps survey.

Johnson and team say that with the addition of UnReview, many existing features within the Create stage will be enriched with machine learning capabilities to speed up the software development lifecycle.

The merge request reviewers feature will be accelerated from a primarily manual process to an automated process using UnReview’s novel machine learning algorithm, which will also be extended in the future to automate other workflow tasks such as the triage of epics and issues including the assigning of issues and suggesting related issues and epics. 

Within the Manage and Plan stages, the UnReview technology provides an improved experience with more intelligent machine learning backed features to automate portfolio management.

The UnReview acquisition leads with business value first and provides GitLab with centralized expertise to build data science workload needs into the entire open DevOps platform.

The UnReview technology will be integrated into the GitLab Code Review ‘experience’ (the organisation’s term, not ours) for GitLab SaaS customers by the end of year. 

Data Center
Data Management