About the Role
In this role you will work closely with deep learning compiler engineers to own and evolve the CI/CD infrastructure that powers the development lifecycle of NVIDIA's deep learning compiler stacks. Responsibilities include designing and operating scalable CI systems that orchestrate ML workloads across diverse GPU and accelerator environments, deliver reliable correctness and performance signals, and serve as a primary technical point of contact for CI health, new project onboarding, and new architecture bring-up.
What You'll Be Doing
•Build, maintain, and improve CI infrastructure that supports development, verification, and release of NVIDIA's deep learning compiler stacks across GPU and accelerator environments
•Improve CI reliability and signal quality by reducing flakes, improving reproducibility, strengthening diagnostics, and making correctness and performance failures easier to understand and act on
•Apply automation, AI, and agent-based workflows to reduce manual CI operations, speed up failure triage, and improve developer efficiency
•Build reusable and self-service CI platforms that support multiple products, projects, model suites, hardware targets, and software configurations while partnering closely with compiler, infrastructure, and release teams
What We Need to See
•BS, MS, or PhD (or equivalent experience) in Computer Science, Computer/Electrical Engineering, Mathematics, or a related field
•5+ years of experience designing, scaling, and operating CI/CD, build/release, or developer infrastructure for complex software systems
•Proven experience building CI platforms end-to-end using systems such as GitLab CI, Jenkins, or similar tools, including pipeline orchestration, compute/runner management, artifact and package systems, and observability, with strong emphasis on reliability, reproducibility, and debuggability
•Strong software engineering skills (Python required), with the ability to design, implement, and debug distributed systems end-to-end
•Familiarity with edge devices (SOC, e.g. NVIDIA Tegra) in host-target architecture, ability to debug it and knowledge of its automation nuances
•Proven track record of designing, building, and deploying AI/LLM-based systems in real engineering workflows, demonstrating skill in evaluating trade-offs, failure modes, maintainability, and measurable impact on developer productivity, signal quality, or operational efficiency
Ways to Stand Out from the Crowd
•Experience crafting and shipping sophisticated AI/agent-based systems that improve continuous integration or developer efficiency. These systems include intelligent test selection, automated triage and routing, regression localization, autonomous remediation, and developer-assist workflows
•Experience operating CI for DL/GPU software environments, including multi-GPU / multi-node workloads on Slurm, Kubernetes, or cloud platforms
•Familiarity with compiler IRs and infrastructure such as LLVM/MLIR, TensorRT IR, especially in the context of testing, debugging, and validating compiler-driven workloads
