Software Engineer, Compute Infra (AI)

Location: San Francisco or New York

Company

Building a full super-intelligence stack spanning large-scale compute infrastructure, models, and applications.

What you will do

• Build sandbox infrastructure that provides each agent with its own VM for coding agents, bots, reinforcement learning, and other workloads.
• Develop a cloud for agents, including the virtualization layer based on Firecracker and Cloud Hypervisor microVMs.
• Rearchitect the control plane and scheduler to support 1 million concurrent VMs per cluster.
• Build a ublk block device with deduplication, lazy loading, and efficient VM data snapshots.
• Design an egress proxy and network isolation to securely filter, monitor, and route agent traffic.
• Design resilient systems operating across cloud and on-premise datacenters.

Requirements

• Bachelor's degree in computer science, engineering, math, or a scientific discipline, or 2+ years of professional experience building software in lieu of a degree.

Nice to have

• Deep experience building compute orchestration systems such as Kubernetes or Mesos.
• Experience building and operating large fleets of VMs or isolated sandboxes.
• Experience with the Linux kernel and virtualization technologies.
• Interest in infrastructure primitives required by future agents.

Culture & Benefits

• Work on infrastructure supporting coding agents, bots, reinforcement learning, and other AI workloads.
• Operate large-scale sandbox clusters spanning third-party and first-party datacenters.