Tech Lead Manager, GPU Cluster Infrastructure

• In frontier AI research, working out the infrastructure is often part of the science. You'd work directly with researchers and other engineers to keep our large-scale experiments performant and fault-tolerant.
• Set the platform's technical direction and own its roadmap. You decide which systems we run, how we schedule and store across providers, and what we measure, from utilization and queue wait to failure rates.
• Stay in the work. You own architecture and the scheduling and storage design, and you debug the failures that cross layers, such as node health, GPU and fabric faults, and multi-node job hangs.
• Hire and grow a small team of senior engineers. You set priorities and ownership, scope projects, and give regular feedback and coaching.
• Set security direction for a shared cluster where AI agents run experiments, covering identity and access, workload isolation, and sandboxing.
• Decide how we operate, from the on-call rotation and incident response to postmortems, fault tolerance, and observability, and take part in it alongside the team.
• Be the escalation point for research teams and for providers, and turn recurring problems into platform fixes.
• You've led engineers as a manager, tech lead, or project lead, setting technical direction, scoping work, and giving feedback. You want to manage people directly; prior direct reports aren't required.
• You've run production Kubernetes for GPU workloads with a batch layer on top (Slurm, Kueue, Volcano, or similar), including quotas, priority and preemption, and node health.
• You've owned infrastructure as code and observability for a production fleet, provisioning with Terraform or Ansible, deploying with Helm and ArgoCD, and monitoring with Prometheus, or their equivalents.
• You're a strong programmer in at least one language that infrastructure is commonly written in, such as Python, Go, Rust, or C++, and your automation and services are maintained as shared code.
• You write clearly for engineers, researchers, and providers, whether it's a roadmap, a design doc, an incident summary, or an escalation.
• If you meet most of this and not all of it, we encourage you to apply anyway.
• Real depth in one or more of these makes you a compelling candidate:
• Distributed training infrastructure: multi-node PyTorch and NCCL debugging, the NVIDIA node stack (drivers, GPU Operator, DCGM), InfiniBand or RoCE fabrics, topology-aware placement.
• Distributed storage: VAST, Weka, Lustre, Ceph, or object storage at scale; checkpoint I/O.
• Cluster security: admission control, RBAC, node and container hardening, sandboxed runtimes (gVisor, Kata, Firecracker), and isolating autonomous agents on shared infrastructure.
• Scheduler internals: Kubernetes scheduler plugins or custom controllers, gang scheduling, fair-share and quota, and the utilization, fairness, and latency tradeoffs between them.
• Multi-provider platforms: scheduling and storage across clusters at different providers so users see one system, including clusters with no shared network and uneven data locality.
• Greenfield team-building: you've hired senior engineers and stood up on-call, incident, and review practices for a new team before.
• Paid Leave - Paid Bereavement, Family, Medical and Pregnancy Disability Leave
• WFH Stipend & Equipment - Work computer and stipend provided for eligible employees
• Catered Meals (Berkeley Office Only) - Catered lunches and dinners on workdays at our office
• (These benefits are currently only offered to full-time US employees, we use an EOR for our international employees who have their own set of benefits for Health, Paid Leave, etc.)

None flagged.