Member of Technical Staff, Sandbox Infrastructure

SF Tensor Company; San Francisco (CA); On-site; USD 150,000 - 190,000; Full time

Benefits offered by this job

Relocation to SF

Job summary

SF Tensor is seeking a highly capable systems engineer to own a serverless GPU container service spanning NVIDIA, AMD, TPU and Trainium. You will extend sandboxing across vendors, implement GPU virtualization, and ensure isolation fidelity for live workloads.

Candidates should have a strong background in Linux kernel, containers, and hypervisors, plus experience with GPU systems, preemption handling, and cross-node sandboxes. Relocation to San Francisco is offered.

Qualifications

• A strong background in low level systems with Linux kernel, containers, namespaces and cgroups.
• Experience with GPU systems, drivers, runtimes or accelerator fleet scheduling.
• Familiarity with distributed failure modes such as preemption, partial failure, checkpoint and restore.
• Fluency in Go, C/C++ or Rust; ability to build missing tools when vendors don’t provide them.

Responsibilities

• Extend the sandboxing stack to new vendors and accelerators.
• Build and maintain GPU virtualization, including gVisor work at driver and ioctl level.
• Ensure sandboxes survive spot capacity with preemption-aware scheduling, checkpointing and rescheduling.
• Support multi-GPU and multi-node sandboxes across NVLink, NVSwitch, InfiniBand and RoCE.
• Own live migration, including socket-preserving migration.
• Guarantee measurement and profiling fidelity under isolation.
• Keep the compiler, post-training and kernel teams from being throughput-limited by sandboxes.

Skills

Low-level systems

Linux kernel

Containers

Namespaces

cgroups

syscall interception

Hypervisors

GPU systems experience

Distributed failure handling

Go

C/C++

Rust

Tools

gVisor

QEMU/KVM

Kata

Job description

Our GPU compiler searches a very wide space of kernel optimizations and verifies the result at the end. That search only works if we can run large volumes of untrusted, newly generated kernels on real hardware, quickly, and without putting anything else at risk.

You will own that layer: a serverless GPU container service spanning NVIDIA, AMD, TPU and Trainium. Its measurements feed the compiler's reward signal, so the sandbox has to be accurate as well as safe. It also carries post-training runs and customer workloads that need hard isolation.

What you'll own

• Extend the sandboxing stack to new vendors and accelerators.
• Build and maintain GPU virtualization, including gVisor work at the driver and ioctl level.
• Make sandboxes survive spot capacity through preemption-aware scheduling, checkpointing and rescheduling.
• Support multi-GPU and multi-node sandboxes across NVLink, NVSwitch, InfiniBand and RoCE.
• Own live migration, including socket-preserving migration.
• Guarantee measurement and profiling fidelity under isolation.
• Keep the compiler, post-training and kernel teams from ever being throughput-limited by sandboxes.

What we're looking for

• A strong low-level systems background: Linux kernel, containers, namespaces, cgroups, syscall interception or hypervisors.
• GPU systems experience across drivers, runtimes or accelerator fleet scheduling.
• Comfort with distributed failure modes: preemption, partial failure, checkpoint and restore.
• Fluency in Go, C/C++ or Rust.
• A strong instinct to build the missing tool when no vendor ships one.

Nice to have

• gVisor, Firecracker, Kata or QEMU/KVM.
• CRIU, live migration or connection-preserving failover.
• NCCL/RCCL, RDMA, InfiniBand or vendor interconnects.
• Running large spot or preemptible fleets.
• Bare-metal provisioning, hypervisor or fleet management.
• Security work on isolation boundaries and untrusted code execution.

How we hire

01

Intro call

30 minutes with a founder. What you have built, and what you want to build next.

02

Systems deep dive

90 minutes on a system you know cold. No puzzles, no whiteboard trivia.

03

Work session

Half a day with the team on a real sandbox problem. Paid.

04

Offer

References, then an offer within a week. Relocation support to San Francisco.