Nebius
System Engineer (Token Factory)
Austin, Texas, United States +6 · Remote · Full-time · Posted Jan 13, 2026
- Location
- Austin, Texas, United States
Remote
Remote - Europe
Netherlands
Remote
United Kingdom
United States - Visa sponsorship
- Applicants must be authorized to work in the country in which they apply and provide proof of employment eligibility as a condition of hire.
About the job
Develop and optimize low-level software and inference runtimes to improve AI model performance and reliability across GPU platforms.
What you'll do
- Develop and optimize low-level kernels and runtime components for AI inference.
- Profile and debug system- and hardware-level performance issues; optimize inference engines on GPU platforms.
- Integrate support for new hardware architectures and collaborate with ML and backend teams on end-to-end execution.
What they ask for
- Strong proficiency in C++ or expertise in GPU programming focused on low-level performance and memory management.
- Experience in GPU programming or systems-level software development.
- Hands-on experience with CPU and GPU profiling and debugging tools, and code optimization based on findings.
- Solid understanding of CPU/GPU architecture and memory hierarchy.
Nice to have
- Experience with GPU computing programming such as CUDA, ROCm, CUTLASS, Cute, ThunderKittens, Triton, Pallas, or Mosaic GPU; familiarity with ML inference runtimes or inference engines.
- Knowledge of Linux internals, drivers, or compiler toolchains; experience with tools such as perf, VTune, Nsight, or ROCm profiler.
Tools
C++, CUDA, ROCm, CUTLASS, Triton, TensorRT, TVM, vLLM
AI focus
Build and optimize an inference platform for foundation models, including text, vision, audio, and multimodal models.
Model-written summary · the original posting is authoritative