AI for CI - Speeding Up Pipelines w/PTS

Maintaining a shared CI infrastructure at scale is no small task. Agent backlogging, redundant job scheduling, and failing builds can collectively become a support nightmare for the engineers who support this automation. Creating a successful CI infrastructure involves optimizing multiple stages of the pipeline, including the test stage.

Predictive Test Selection is a new method for performing test impact analysis, which can safely reduce the number of tests that need to be executed per build cycle. It uses machine learning to analyze a history of test behavior and predict which tests are likely to produce valuable feedback in a given build context. Many complex builds spend the majority of their time executing tests, so reducing the amount of testing work that needs to be accomplished can have a positive impact on CI agent utilization.

In this webinar, Justin Reock, Field CTO at Gradle Enterprise, will explain how Predictive Test Selection works, where it fits into the overall CI workflow, and how organizations are already saving time and resources by utilizing this new technique.