Worked on the EnzymeAD/Enzyme repository to streamline continuous integration workflows and improve data handling practices. Focused on removing automatic benchmark results uploading, which reduced external data exposure and simplified the CI pipeline. Enhanced reliability and efficiency by updating CI runners to newer Ubuntu versions, optimizing trigger conditions for Fortran builds, and introducing setup-bazel caching. Leveraged skills in Bazel, GitHub Actions, and scripting with Python and YAML to deliver features that accelerated developer feedback and reduced maintenance overhead. The work resulted in faster, more reliable build cycles and safer data management, directly supporting improved developer velocity and more robust release processes.