During May 2026, work centered on developing a robust build script for the ml-commons 3.5.0 stack within the ppc64le/build-scripts repository. The approach emphasized reliable build automation and precise dependency management, using Java and Bash to ensure reproducible builds and streamline release cycles. By aligning the build workflow with CI/CD practices, the project improved onboarding and maintenance for downstream machine learning workloads on PPC64LE environments. The implementation featured effective version pinning and enhanced repository hygiene, reducing version drift and simplifying future upgrades. This work demonstrated strong skills in script writing, build tooling, and managing complex dependencies in large-scale environments.