Research Scientist / Engineer – Performance Optimization
About this role
Research Scientist / Engineer – Performance Optimization is a full-time role based in Redwood City, CA at Luma AI that suits professionals with expert GPU-optimization skills and deep transformer knowledge. The position centers on making multimodal models fast through profiling and optimizing hardware utilization for training and deployment at scale.
Role Overview
This position involves deep performance work on multimodal models at Luma AI. The professional selected for this job will write custom operations, fused kernels, and optimize tensor cores across hardware platforms to achieve minimal latency and efficient scaling.
Key Responsibilities
Requirements
Nice to Have
Skills and Tools
Salary and Benefits
About the Company
Luma AI focuses on building unified general intelligence capable of generating, understanding, and operating within the physical world. The organization emphasizes that multimodality is essential for intelligence, moving beyond language models by incorporating vision. Luma AI operates as an equal opportunity employer.
How to Apply
Interested candidates should apply on the official company careers page.
Skills and tech stack
Questions candidates ask
What is the location of the Research Scientist / Engineer – Performance Optimization role?
The position is located in Redwood City, CA.
What salary range is offered for this role?
The salary ranges from $195K to $395K.
What core technical skills are required?
Requirements include expert-level Triton and CUDA programming, GPU optimization, strong PyTorch skills including custom operations, proficiency with profiling tools like NVIDIA Nsight and torch profiler, and a deep understanding of transformer architectures and attention mechanisms.
What are some of the nice-to-have qualifications?
Preferred additions include experience with compilers and exporters such as torch.compile, TensorRT, ONNX, and XLA, inference workload optimization, Triton compiler and kernel fusion techniques, and knowledge of warp-level intrinsics and advanced CUDA optimization.
What is the mission of Luma AI?
Luma AI aims to build unified general intelligence that generates, understands, and operates in the physical world through vision and multimodality.