Make AI accelerator hardware feel native inside Kubernetes clusters.
As a Staff Cloud Software Engineer, you will build Go services that connect accelerator hardware with Kubernetes. You will own software components that turn complex compute capability into a usable platform experience.
Your work will shape how AI workloads discover, access and scale across cluster infrastructure. That means designing production services where reliability, performance and observability are part of the architecture.
You will build CRDs, operators and Kubernetes components for ML and platform teams. You will partner with hardware, cloud and ML engineers on design choices and roadmap priorities.
This is a senior IC role with room to set direction. You will mentor engineers, raise code quality and help define how Graphcore delivers cloud-native AI compute.
The Workload Management team sits between Kubernetes, cloud platforms and accelerator hardware. Work moves from architecture to implementation, review and production readiness without losing system context.
You will be trusted to own technical areas, challenge assumptions and move important work forward. The pace is driven by clear priorities, fast feedback and strong engineering judgement.
Decisions are made through evidence, prototypes and direct technical discussion. Engineers are expected to speak up early, share trade-offs and help the team learn quickly.