GPU Performance Engineer

What You’ll Do:

The GPU Performance Team is responsible for the automated testing and verification of CoreWeave’s ever-expanding fleet of nodes and infrastructure.   The team plays a critical role in ensuring that the hardware we deliver to customers is stable and performant throughout its lifecycle.  

About the role:

We are seeking a Senior Software Engineer to join the GPU Performance team to help us build and refine our automation platforms and workflows for reliability, efficiency, scalability and observability. This individual will join a team of Golang and Python developers innovating kubernetes-based solutions to hardware performance testing. As a member of the GPU Performance  team, you would have the opportunity to:

• Design and Implement solutions for testing infrastructure at scale through automation.
• Develop platforms, workflows, data structures, API endpoints, and other testing management tools and capabilities.
• Create test plans, deployment automation, dashboards, alerts, and insights into our performance testing platform.
• Grow, change, invest in your teammates, be invested-in, share your ideas, listen to others, be curious, have fun, and, above all, be yourself.

Who You Are: 

You are a software engineer with a strong bias toward reliability, automation, and scale. You’re comfortable owning services end‑to‑end and working close to hardware, but your primary tools are Go or Python, Kubernetes, and well‑designed APIs.

• Have experience building and maintaining backend services and APIs (REST/gRPC) and are comfortable integrating with infrastructure systems (Kubernetes, CI/CD, telemetry, metrics backends).
• Are motivated by automation at scale: you’d rather design a controller, workflow, or platform than run one‑off tests by hand.
• Are familiar with system metrics, performance, and health (e.g., GPU/CPU, network, storage, or similar infra) and know how to turn raw metrics into dashboards, alerts, and actionable insights.
• Are comfortable designing and implementing performance tests and validation workflows that run across many nodes and clusters.
• Care about code quality, observability, and operational readiness: logs, metrics, tracing, alerting, and on‑call are part of how you think about building systems.
• Are willing to participate in an on‑call rotation and help own the stability of the platforms you build.
• Work well in a collaborative, curious, and feedback‑driven team: you invest in your teammates, share ideas openly, and are excited to help shape how CoreWeave validates hardware performance at fleet scale.

Preferred:

• Write and maintain Kubernetes custom controllers and operators to automate infrastructure testing.
• Contributions to OSS projects such as  llm-d, vLLM, or PyTorch
• Exposure to benchmarking large GPU fleets or multi-region clusters.
• Experience with CUDA kernels