Staff ML Engineer, Optimization

We are seeking a Senior ML Infrastructure Engineer to join our team at Runway. In this role, you will be responsible for optimizing our AI models for media generation, specifically focusing on video generation. We are looking for someone with extensive experience in improving model training and inference performance, as well as a strong understanding of GPU workloads. The ideal candidate will thrive in a fast-paced, high-ownership environment and have a background in Python, C/C++, CUDA, and profiling GPU performance. Strong communication and collaboration skills are also essential for this role.

• Optimize state-of-the-art AI models for video generation, such as Gen-1 and Gen-2
• Build tooling to improve the efficiency and reliability of distributed training runs on Runway's HPC cluster
• Have 3+ years of experience in a role optimizing machine learning model inference and training on NVIDIA hardware
• Have knowledge of Python, C/C++, CUDA, and extensive experience profiling GPU performance and distributed training runs
• Make contributions to an ML framework (such as PyTorch), optimized runtimes for inference (such as TensorRT) or compilers (such as GCC)
• Possess strong communication, collaboration, and documentation skills

• 3+ years of experience in a role optimizing machine learning model inference and training on NVIDIA hardware
• Knowledge of Python, C/C++, CUDA, and extensive experience profiling GPU performance and distributed training runs
• Contributions to an ML framework (such as PyTorch), optimized runtimes for inference (such as TensorRT) or compilers (such as GCC)
• Strong communication, collaboration, and documentation skills

• Competitive market rates for salary
• Overall compensation package
• Consideration of candidates with varying levels of experience
• Communication of any updates in expected salary range
• Equal opportunity for success regardless of race, gender identity or expression, sexual orientation, religion, origin, ability, age, veteran status
• Recognition as a best place to work