Luma AI
Research Scientist / Engineer – Performance Optimization
Redwood City, CA +5 · Hybrid · Full-time · Posted Oct 8, 2026
- Location
- Redwood City, CA
Remote, US
Remote, EU
New York, NY
Seattle, WA
London, UK
About the job
Optimize Luma’s multimodal models by profiling and improving hardware performance, developing high-performance kernels, and preparing models for distributed production deployment.
What you'll do
- Profile and optimize GPU, CPU, and accelerator code for utilization and latency.
- Develop high-performance PyTorch, Triton, and CUDA kernels and custom operations.
- Optimize transformer models and implementations for distributed deployment, and build performance monitoring and automation.
What they ask for
- Expert-level Triton/CUDA programming and GPU optimization.
- Strong PyTorch skills, including kernel development and custom operations.
- Proficiency with profiling tools such as NVIDIA Nsight and torch profiler.
- Deep understanding of transformer architectures and attention mechanisms.
Nice to have
- Experience with compilers and exporters such as torch.compile, TensorRT, ONNX, or XLA.
- Experience optimizing inference workloads for latency and throughput.
Tools
PyTorch, Triton, CUDA, NVIDIA Nsight, torch profiler, torch.compile, TensorRT, ONNX
AI focus
Research and implement optimization techniques for transformer models and optimize Luma’s multimodal models for efficient training and deployment.
Model-written summary · the original posting is authoritative