Staff Engineer, Serverless GPU Sandbox & Live Migration

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

Benefits offered by this job

Relocation assistance

Equity

Competitive compensation

Job summary

SF Tensor in San Francisco is hiring for a Member of Technical Staff for Sandbox Infrastructure to build a serverless GPU container service across multiple vendors. You will extend sandboxes, ensure isolation and measurement fidelity, and enable live migration with socket preservation.

This is a high-performance compute role requiring deep systems expertise. You’ll work with Go, C/C++ or Rust, tackle preemption, and scale sandboxes across GPUs and nodes while collaborating with compiler and

Qualifications

• Strong low-level systems engineering background.
• Experience with GPU drivers, runtimes or scheduling.
• Familiarity with containers, namespaces, cgroups and hypervisors.
• Proficiency in Go, C/C++ or Rust.

Responsibilities

• Extend sandboxing stack to new vendors and accelerators.
• Build and maintain GPU virtualization below the runtime.
• Support spot capacity with preemption-aware scheduling and rescheduling.
• Enable multi-GPU and multi-node sandboxes with NVLink, InfiniBand paths.
• Own live migration end-to-end with socket-preserving migration.
• Ensure measurement and profiling fidelity and isolation.
• Collaborate with compiler, post-training and kernel teams.

Skills

Linux kernel

Go

C/C++

Rust

Distributed systems

GPU systems

Job description

SF Tensor in San Francisco is hiring for a Member of Technical Staff for Sandbox Infrastructure to build a serverless GPU container service across multiple vendors. You will extend sandboxes, ensure isolation and measurement fidelity, and enable live migration with socket preservation.

This is a high-performance compute role requiring deep systems expertise. You’ll work with Go, C/C++ or Rust, tackle preemption, and scale sandboxes across GPUs and nodes while collaborating with compiler and