Online
Why did my build fail? Using AI to troubleshoot faster with Failure Summaries
Most developers tell us that working with CI is frustrating in large projects. Making a change in large projects can generate hundreds of builds with various types of failures—not to mention overall background flakiness. When something goes wrong, troubleshooting issues requires manually navigating several UIs and many large log files.
This is where "Failure Summaries", an upcoming Develocity feature, comes in to save the day. Built in-house for Gradle engineers, Failure Summaries use AI for categorization and clustering to determine which tests are primarily responsible for a number of failures.
In this session with Laurent Ploix, Senior Product Manager at Gradle, you'll see how Failure Summaries can streamline troubleshooting with a deep dive into how to distill complex test failures.
- What measuring developer productivity looks like, which metrics matter, and how we want to simplify troubleshooting
- The effectiveness of extracting CI feedback through the lens of specific evaluation metrics, for example, error log segments, failure clustering into logical groups, root cause identification, and pinpointing infrastructure issues
- Concrete examples are seen in action in the Gradle team's own CI environment.
Meet your speaker
Sponsored by
Past events
Online
Software Delivery Excellence in the age of AI: A Fireside Chat with Martin Odersky and Hans Dockter
From build tooling to language design, we'll explore how we can stay productive as the JVM world absorbs the new wave of AI complexity.
FREE
Online
Testing software is awful: Here is how we can fix it
Join Java Champions Brian Demers and Trisha Gee as they cover top challenges developers face when testing software, like dealing with long test cycles, slow and frustrating troubleshooting, a "forever war" with flaky tests, and the upcoming impact of AI-generated tests on projects.
FREE