Software Development · San Francisco, CA

On-site Full-time

Principal Software Engineer

Crusoe · San Francisco, CA · Posted Nov 6, 2026

This posting comes from Crusoe. Confirm the details on their site before applying.

Overview

As a Principal Software Engineer for Infrastructure, you will partner with teams across the company to help shape the technical direction of the compute, storage, and networking platforms that power Crusoe’s AI-first cloud. This is a high-leverage leadership role that spans the full infrastructure stack: you will work alongside engineers and stakeholders on the systems that help massive GPU clusters run AI and HPC workloads at extreme throughput, low latency, and high reliability.

Responsibilities

  • Infrastructure Strategy: Partner with engineering, product, and leadership to help shape the long-term direction for compute, storage, and networking, and to turn that direction into practical plans and milestones.
  • Architecture & Design: Collaborate with teams on architecture reviews, design documents, and code reviews for critical infrastructure components and services across all three domains.
  • Compute & Virtualization: Work with compute teams on the virtualization and orchestration layers that deliver bare-metal-class performance to GPU workloads, including hypervisor tuning, SR-IOV, vDPA, virtio, device passthrough, and NUMA- and topology-aware scheduling.
  • Storage Systems: Work with storage teams on high-throughput, low-latency storage for training and inference pipelines, including distributed block, file, and object storage, NVMe and NVMe-oF, RDMA-based data paths, and checkpointing and data-loading performance.
  • Networking & Fabric: Partner with networking teams on architectures for GPU-to-GPU fabrics, high-speed Ethernet, and specialized AI clusters, using technologies such as XDP, eBPF, DPDK, OVS/OVN, BGP, EVPN, and VXLAN.
  • Kernel & Driver Work: Contribute to Linux kernel modules and high-performance drivers across the network, storage, and device stack.
  • Hardware Integration: Work with Hardware teams on accelerators and offload devices, including GPUs, SmartNICs, DPUs, IPUs, and NVMe, from qualification through bring-up of new platforms.
  • Cross-Stack Performance & Reliability: Team up with on-call, SRE, and domain engineers to troubleshoot complex production issues that cross compute, storage, and network boundaries, and to improve performance, observability, capacity planning, and failure-domain design.
  • Technology Evaluation: Explore emerging infrastructure technologies with the team, contribute to open-source projects where strategically relevant, and share learnings with the broader organization.
  • Mentorship & Collaboration: Partner with internal stakeholders on end-to-end features, and mentor and support engineers across teams to foster a culture of engineering excellence.

Requirements

  • Large-Scale Infrastructure Mastery: 10+ years of experience building and operating large-scale infrastructure systems in production, ideally within cloud or hyperscale data center environments.
  • Cross-Domain Breadth: Hands-on experience across compute, storage, and networking, and a track record of solving problems that cross domain boundaries.
  • Systems Programming Excellence: Strong systems programming skills in C, C++, Rust, or Go, with experience working on performance-critical services and control planes.
  • Linux Internals Depth: Extensive knowledge of the Linux kernel architecture, memory management, scheduler, I/O and network stacks, and device driver development.
  • Virtualization & Compute Literacy: Experience with KVM/QEMU or similar hypervisors, container runtimes, and orchestration platforms such as Kubernetes, plus an understanding of GPU, CPU, and NUMA topology and its effect on workload performance.
  • Storage Literacy: Familiarity with distributed storage architectures (block, file, or object), NVMe/NVMe-oF, RDMA, and the consistency, durability, and performance tradeoffs involved.
  • Network & Fabric Literacy: Familiarity with data center fabrics (e.g., Spine-Leaf, BGP, EVPN, VXLAN), network virtualization tools such as OVS or OVN, and high-speed packet processing technologies such as XDP, eBPF, or DPDK.
  • Accelerator Knowledge: Familiarity with SmartNICs, DPUs, and accelerators, such as NVIDIA ConnectX-6/7 and BlueField-3 or Intel IPUs.
  • Production Debugging: Experience working with others to diagnose complex issues across the full stack, from hardware and firmware through the kernel to control-plane software.
  • Strategic Influence: A proven track record of leading through influence, with excellent written communication skills and a strong mentoring mindset.
  • Education: A Bachelor’s degree in Computer Science, Computer Engineering, or a related analytical field (or equivalent professional experience).

Preferred qualifications

  • Open-Source Contributions: Active contributions to projects such as the Linux kernel, KVM/QEMU, Kubernetes, Ceph, OVS, OVN, Cilium, or Multus.
  • Orchestration & Automation Expertise: Experience with infrastructure automation, fleet management, or complex orchestration tooling.
  • GPU Training at Scale: Prior experience building compute, storage, or networking stacks for multi-node GPU training at scale.
  • Hardware Co-Design: Experience working with hardware teams on platform design, qualification, and firmware-to-software integration.

Benefits

  • Competitive compensation and equity packages
  • Restricted Stock Units
  • Paid time off, paid holidays & leave of absence programs
  • Comprehensive health, dental & vision insurance
  • Employer contributions to HSA account
  • Paid parental leave
  • Paid life insurance, short-term and long-term disability
  • Professional development & tuition reimbursement
  • Mental health & wellness support
  • Commuter benefits (parking & transit)
  • Cell phone stipend
  • 401(k) Retirement plan with company match up to 4% of salary

Great fits start here.

Whether you're looking for your next role or your next team member, give people and possibilities a place to meet.