Global AI MapResearch Scientist / Engineer – Performance Optimization
Redwood City, CA; Remote, US; Remote, EU; New York, NY; Seattle, WA; London, UK
Role
The role involves profiling and optimizing GPU, CPU, and accelerator code for Luma's multimodal models to improve training efficiency and deployment scale. Candidates must have expert-level skills in Triton, CUDA, and PyTorch, along with a deep understanding of transformer internals and hardware features.
Location: Redwood City, CA; Remote, US; Remote, EU; New York, NY; Seattle, WA; London, UK
Employment: unspecified
Work pattern: hybrid
Seniority: senior
APPLY ON THE EMPLOYER’S SITE ↗
Source and currency
This vacancy came from the organization’s official Ashby feed. Source update: 2026-10-08T15:06:04.726+00:00. Check the employer’s page for the complete description, current availability and application terms.