NVIDIA is hiring a Senior CUDA Driver and DevOps Engineer to lead the decomposition of complex build processes and ensure robust CI/CD pipelines. In this role, you will automate infrastructure across multiple platforms and environments, championing best practices and collaborating with engineering leadership.

What You’ll Do

Decompose and modularize build processes for reusability across multiple projects
Debug GitHub Actions and GitLab pipelines to ensure timely and efficient CI execution
Script and build infrastructure to handle dependencies across various environments and build systems
Bring up builds and CI across platforms (x64/arm64) and OSes (Linux/Windows/Mac)
Work with engineering leadership to identify the support matrix and define the scope of the build matrix
Craft and update documentation and coordinate with partners on multi-functional projects
Automate scheduled work for all of the above

What We’re Looking For

Bachelor’s Degree in Systems/Software/Computer Engineering, CS or equivalent experience
8+ years of relevant industry experience or equivalent academic experience after BS
Experience working across multiple highly-coupled projects in a VCS like Git
Experience collaborating with cloud providers, Kubernetes, GitHub Actions, and other systems
Familiarity with automating container builds, updates, and debugging multi-container workflows
Background with CI/CD systems including Github and Gitlab
Understanding of testing principles and how to quantify/improve coverage and developer velocity
Knowledge of release management practices
Strong analytical, debugging, and problem-solving skills
Familiarity with containerization technologies (e.g., Docker)

Nice to Have

Experience working with or building for HPC/multi-node environments
Experience handling proprietary software, confidential hardware, or large codebases (100k+ LoC)
Familiarity with deployment technologies for getting software from CI/CD to customers
Exposure to development across multiple OSes
Familiarity with build systems such as CMake, autotools, make, bazel or others