Senior Software Engineer, Sandboxes & Virtualization

Overview In this role you design and evolve secure, sandboxed runtime environments for Kubernetes-based workloads, focusing on isolation, performance, and security. You will build platforms that blend container runtimes, lightweight VMs, and virtualization to support GPU-accelerated workloads in multi-tenant setups, addressing latency-sensitive needs. You’ll influence Linux internals, runtimes, virtualization layers, and GPU drivers for scalable, reliable operations. This is a chance to shape high-performance AI infrastructure at scale with a collaborative, fast-paced team at CoreWeave. Compensation / BenefitsMedical, dental, and vision insurance401(k) with employer matchFlexible PTOPaid Parental LeaveTuition ReimbursementEmployee Stock Purchase Program ResponsibilitiesDesign secure sandboxed runtime environments for Kubernetes workloadsDevelop platforms integrating container runtimes, lightweight VMs, and virtualizationEngineer support for GPU-accelerated workloads in multi-tenant environmentsCreate GPU-aware sandboxing and scheduling strategiesOptimize containerization and I/O for latency-sensitive workloadsBuild systems to safely execute untrusted or shared GPU workloadsContribute to architecture decisions across Linux internals, container runtimes, virtualization layers, and GPU drivers Key requirements3+ years in systems/platform/infrastructure engineeringStrong hands-on Kubernetes and cloud-native architecturesExperience designing/operating secure execution environmentsPractical experience with lightweight virtualization/sandboxing (e.g., Kata Containers, gVisor, KubeVirt, QEMU)Experience with GPU-accelerated workloads in multi-tenant environmentsProficiency in systems programming (Go, C/C++, Rust, Bash) and Linux internalsExperience diagnosing performance/reliability/isolation issues across containers, VMs, and infraExperience profiling, benchmarking, and performance tuning across runtime, virtualization, and GPU stacksCuriosity about security, virtualization, Kubernetes, and GPU infrastructureOwnership mindset and proactive problem-solvingAbility to reason about tradeoffs between isolation, performance, and operabilityKubernetes and container orchestrationContainer runtimes and sandboxing technologies (Kata Containers, gVisor, KubeVirt, QEMU)GPU scheduling, device passthrough, mediated devices, GPU virtualization concepts