Member of Technical Staff - GPU Infrastructure

About the role

Hyperbolic Labs is building an open‑access AI cloud that aggregates idle GPU resources worldwide. We need a senior infrastructure engineer to design and operate the core provisioning and virtualization layer that turns raw GPUs into a multi‑tenant, programmable pool for thousands of AI developers.

Key responsibilities

• Design and implement a multi‑tenancy provisioning system for diverse GPU hardware.
• Develop automated bare‑metal lifecycle workflows using IPMI/Redfish, PXE boot, and cloud‑init.
• Build GPU scheduling and orchestration logic that accounts for type, memory, topology, and fragmentation.
• Create and maintain infrastructure as code with Terraform or Pulumi, including CI/CD pipelines, secrets, and observability.
• Integrate storage solutions (object storage, high‑IOPS block, distributed file systems) for AI/ML training data and checkpoints.
• Collaborate with hardware vendors to troubleshoot and optimise hardware integrations.

Required profile

• Deep expertise in bare‑metal provisioning, remote management (IPMI/Redfish), and automated OS deployment.
• Strong knowledge of GPU architecture, CUDA, and compute optimisation.
• Proven experience building and scaling cloud infrastructure or distributed systems in production.
• Excellent communication skills and ability to work with cross‑functional teams.

Required skills

• IPMI, Redfish, BMC remote management
• PXE boot, cloud‑init, automated OS deployment
• GPU scheduling, topology awareness, fragmentation minimisation
• Terraform or Pulumi
• CI/CD for infrastructure, secrets management, observability stack
• Object storage, high‑IOPS block storage, distributed file systems (e.g., Ceph, Weka, VAST Data)
• API design
• CUDA and GPU compute optimisation
• InfiniBand, RoCE (RDMA) networking

What we offer

• Opportunity to shape a pioneering AI cloud platform.
• Collaborative, open‑source‑focused engineering culture.
• Competitive compensation and equity.