Staff Engineer, GPU Sandbox Infrastructure

San Francisco Tensor Company; San Francisco (CA); On-site; USD 275,000 - 315,000; Full time

Benefits offered by this job

Relocation assistance

Meaningful equity

Office in San Francisco

Job summary

San Francisco Tensor Company is hiring a Member of Technical Staff for Sandbox Infrastructure to build a serverless GPU container service across NVIDIA, AMD, TPU and Trainium, scalable and isolated for untrusted code.

You will work with compiler, post‑training and kernel teams to ensure throughput remains uncapped by sandboxes while maintaining fidelity and security. Relocation to San Francisco is supported where possible.

Qualifications

• Strong low-level systems engineering background: Linux kernel internals, containers, namespaces, cgroups, syscall interception or hypervisors.
• Experience in GPU systems engineering: drivers, runtimes or scheduling on accelerator fleets.
• Familiar with distributed systems failure modes: preemption, partial failure, checkpoint/restore and dealing with states you can't afford to loose.
• Proficient in Go, C/C++ or Rust.
• Bias toward building the thing yourself when no vendor supports what you need.

Responsibilities

• Extend our sandboxing stack to new vendors and accelerators.
• Build and maintain GPU virtualization below the runtime, including gVisor work at the driver and ioctl level.
• Make sandboxes first‑class citizens on spot capacity, with preemption‑aware scheduling, checkpointing and rescheduling.
• Support multi‑GPU and multi‑node sandboxes, including interconnects (NVLink, NVSwitch) and RDMA paths (InfiniBand, RoCE).
• Own live migration end‑to‑end, including socket‑preserving migration.
• Guarantee measurement and profiling fidelity as well as their isolation.
• Work directly with compiler, post‑training and kernel teams to ensure throughput is not capped by sandboxes.

Skills

Go

C/C++

Rust

Linux kernel internals

Containers

GPU systems engineering

Tools

gVisor

Firecracker

Kata

QEMU/KVM

CRIU

NCCL/RCCL

RDMA

InfiniBand

Job description

San Francisco Tensor Company is hiring a Member of Technical Staff for Sandbox Infrastructure to build a serverless GPU container service across NVIDIA, AMD, TPU and Trainium, scalable and isolated for untrusted code.

You will work with compiler, post‑training and kernel teams to ensure throughput remains uncapped by sandboxes while maintaining fidelity and security. Relocation to San Francisco is supported where possible.