Senior Software Engineer, Compute Platform

Role overview Develop the compute orchestration platform behind GPU-accelerated training, inference, simulations, and research workloads. The work includes provisioning bare-metal servers, scheduling compute resources, managing GPU clusters, and building interfaces that let users access high-performance infrastructure programmatically.

Responsibilities - Build orchestration services for GPU clusters, bare-metal provisioning, and resource allocation across data center environments. - Develop scheduling and optimization approaches that improve GPU utilization while meeting workload performance needs. - Automate the provisioning and ongoing lifecycle management of research clusters, including Kubernetes and SLURM environments. - Create platform capabilities for GPU configuration, workload placement, and tenant isolation. - Optimize platform components for high throughput and low latency, including in containerized or virtualized environments. - Build APIs, SDKs, monitoring, and telemetry for compute resource management and infrastructure visibility.

Requirements - At least five years of software engineering experience building production compute platforms, orchestration systems, or distributed compute infrastructure. - Strong background in compute orchestration, resource scheduling, or workload management at scale. - Familiarity with Kubernetes architecture and container orchestration. - Experience with Go, C/C++, Python, or Rust, and strong Linux production skills, including performance optimization and low-level resource management. - Knowledge of virtualization technologies, container runtimes, and orchestration platforms. - Understanding of GPU architecture, CUDA or GPU resource management, or the ability to learn these areas quickly.

Nice to have - Experience managing Kubernetes, SLURM, or HPC research clusters; GPU virtualization or multi-tenant sharing; or custom scheduling. - Knowledge of GPU interconnects and high-performance networking such as InfiniBand, RDMA, NVLink, or NVSwitch. - Familiarity with AI training frameworks, distributed training, or infrastructure for research and regulated environments.

Benefits and work setup The listing gives a total compensation range of $165,000–$225,000, including base pay and equity, plus a 6% 401(k) match and fully covered health insurance premiums.