MTS - Research (Kernels & Hardware RL)

What you'll do

• Design executable tasks for GPU and accelerator kernels, low-level systems programming, and hardware-aware optimization.
• Build correctness tests, performance graders, and secure execution sandboxes; analyze agent performance and failure modes.
• Develop compilation, execution, profiling, and evaluation pipelines, and collaborate on post-training experiments.

What they ask for

• Bachelor’s, Master’s, or PhD in CS, Electrical or Computer Engineering, or a related technical field, or demonstrated proof of work through significant open-source contributions or industry experience.
• Strong C/C++ and CUDA or a comparable kernel language, alongside production-quality Python; experience building and profiling kernels.
• Hands-on experience with LLMs and understanding of how they are evaluated and optimized.
• Experience conducting reproducible performance experiments and interpreting profiler results.

Nice to have

• Expertise in Triton, CUTLASS, LLVM/MLIR, GPU architectures, vector ISAs, FPGA/HLS, RTL/HDL, EDA tools, or architecture simulators; experience building RL environments, agent harnesses, or evaluation infrastructure.
• Experience fine-tuning or evaluating large-scale models to deliver “frontier performance” on open-source benchmarks.

Pay

$160,000 to $400,000 per year (base salary)

Tools

C/C++, CUDA, Python, Triton, CUTLASS, LLVM/MLIR, FPGA/HLS, RTL/HDL

AI focus

Build reinforcement learning environments and evaluation stacks for agents that write efficient GPU kernels and solve hardware optimization problems; collaborate on post-training experiments to improve agent capabilities.