Expert Research Scientist in Performance Optimization

Become an Expert Research Scientist in Performance Optimization with Luma AI, committed to transforming multimodal AI capabilities through enhanced model performance and efficiency. This role supports remote and hybrid work.

At Luma AI, you'll be at the forefront of maximizing the performance of AI systems as part of the Performance Optimization team. Utilizing your extensive knowledge in Triton and CUDA, you will work collaboratively to ensure efficient training and deployment of advanced AI models while maintaining the highest quality standards.

Key Responsibilities: • Optimize GPU/CPU/Accelerator code for efficacy • Develop high-performance solutions with PyTorch and Triton • Create fused kernels leveraging state-of-the-art hardware • Enhance model architectures for production settings • Design automated performance monitoring systems

Requirements: • Expert in Triton and CUDA programming • Strong experience with PyTorch • Proficient with GPU profiling tools • Deep knowledge of transformer architecture • Familiarity with compiler optimization techniques

Elevate your career by enhancing AI performance at Luma AI.