Principal Software Engineer
Core
Architect and optimize AI systems across hardware, compilers, kernels, frameworks, runtimes, and distributed infrastructure to improve training and inference performance.
Role type
Principal Software Engineer (AI Systems & Hardware Acceleration)
Builds
High-performance AI training and inference solutions on large accelerator clusters
Domain
AI Infrastructure / Hardware-Software Co-design
Deliverable
production ML models
Required skills
CUDA, ROCm, Triton, C++, C#, Java, Python, distributed training, quantization, MoE, memory management, parallelism strategies, AI frameworks (PyTorch)
Preferred skills
hardware-software co-design, silicon roadmap influence, publications/patents, open-source contributions
Technologies
PyTorch, Triton, vLLM, SGLang, CUDA, ROCm
Responsibilities
Partner with hardware architecture teams on hardware-software co-design; Drive performance optimization across kernels, communication, and memory movement; Architect distributed training and inference solutions; Lead complex technical initiatives across multiple engineering organizations.
Seniority
Principal, strategy & mentorship