Compute Infrastructure Member of Technical Staff

Requirements

• Deep expertise in virtualization technologies (KVM, Xen, QEMU) and advanced containerization/sandboxing (Kata, Firecracker, gVisor, Sysbox or equivalent)
• Strong proficiency in systems programming languages such as C/C++ and Rust
• Proven track record profiling, debugging, and optimizing complex system-level performance issues, with deep knowledge of Linux kernel internals, resource management, scheduling, memory management, and low-level engineering
• Hands‑on experience building or significantly enhancing distributed compute platforms, orchestration systems, or high‑performance infrastructure at scale
• Ability to thrive in a fast‑paced, meritocratic environment with full ownership, high standards, and a focus on rigorous execution
• (Desirable) Experience in Linux kernel development, hypervisor extensions, or low‑level system programming for compute‑intensive workloads
• (Desirable) Proven track record operating or designing large‑scale AI training/inference clusters (GPU/TPU scale)
• (Desirable) Experience with custom runtimes, isolation techniques, or bespoke platforms for specialized AI compute
• (Desirable) Familiarity with performance tools, tracing, and debugging in production distributed environments

• The Compute Infrastructure team at xAI is responsible for designing, building, and operating the massive‑scale clusters and orchestration platforms that power frontier AI training, inference, and agent workloads at unprecedented scale
• In this role, you will push the boundaries of container orchestration far beyond existing systems like Kubernetes, manage exascale compute resources, optimize for high‑performance training runs and production serving, and collaborate closely with research and systems teams to deliver reliable, ultra‑scalable infrastructure that enables xAI's next‑generation models and applications
• Build and manage massive‑scale clusters to host, persist, train, and serve AI workloads with extreme reliability and performance
• Design, develop, and extend an in‑house container orchestration platform that achieves superior scalability, isolation, resource efficiency, and fault‑tolerance compared to off‑the‑shelf solutions
• Collaborate with research teams to architect and optimize compute clusters specifically for large‑scale training runs, inference services, and real‑time applications
• Profile, debug, and resolve complex system‑level performance bottlenecks, resource contention, scheduling issues, and reliability problems across the full stack
• Own end‑to‑end infrastructure initiatives with first‑principles design, rigorous testing, automation, and continuous optimization to support frontier AI compute demands