Worked on the JetBrains/intellij-community repository to enhance the Gradle-based evaluation pipeline by delivering Docker-aware Gradle evaluation and refining the Java test runner. Focused on improving test result calculations and filtering out unnecessary test commands, the work aimed to reduce noise and provide cleaner outputs for developers. Leveraged Kotlin and Gradle to establish a more robust evaluation build workflow, which increased the reliability of the pipeline and enabled faster feedback loops. The technical approach included integrating Docker evaluation support and optimizing test-result computation, resulting in more reliable evaluation metrics and a streamlined experience for backend development and continuous integration processes.