Senior GPU Cloud Infrastructure Engineer

Hyperbolic; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Hyperbolic is seeking a very senior Infrastructure Engineer to scale a GPU cloud marketplace. You will build a multi-tenant provisioning and virtualization layer, turning global GPU inventories into an orchestrated pool for AI developers and researchers.

You will own the core orchestration, drive cost efficiencies, and work closely with hardware vendors to optimize integrations across production cloud infrastructure.

Qualifications

• Extensive experience with bare-metal provisioning and lifecycle management.
• Deep knowledge of GPU scheduling, memory, topology and multi-GPU placement.
• Proficiency with Terraform or Pulumi and infra-focused CI/CD pipelines.
• Experience with AI/ML storage, object and block storage, and FS for training data.
• Strong API design skills and automation using cloud-init for provisioning.
• Solid CUDA and GPU compute optimization understanding.
• Collaborative cross-functional work with hardware vendors and teams.

Responsibilities

• Build and scale multi-tenant provisioning and virtualization for a GPU cloud marketplace.
• Transform raw GPUs from global suppliers into a programmable, orchestrated pool.
• Develop the core orchestration layer to enable cost savings vs traditional clouds.
• Ensure reliability, security, and observability across production infra.
• Collaborate with hardware vendors and internal teams to optimize integrations.

Skills

Bare-metal provisioning

GPU scheduling

Terraform / Pulumi

CI/CD for infra

Observability stack

Storage for AI/ML

API design

CUDA knowledge

Vendor collaboration

Distributed systems

Job description

Hyperbolic is seeking a very senior Infrastructure Engineer to scale a GPU cloud marketplace. You will build a multi-tenant provisioning and virtualization layer, turning global GPU inventories into an orchestrated pool for AI developers and researchers.

You will own the core orchestration, drive cost efficiencies, and work closely with hardware vendors to optimize integrations across production cloud infrastructure.