About the job
At Roblox, we’re building the tools and platform that empower our community to bring any experience that they can imagine to life. Our vision is to reimagine the way people come together, from anywhere in the world, and on any device. We’re on a mission to connect a billion people with optimism and civility, and looking for amazing talent to help us get there. As a Model Optimization engineer on ML Platform, you will be responsible for digging deep into model internals to optimize performance, for both training and inference.
Responsibilities
Optimize machine learning models for performance on GPU architectures, focusing on both training and inference workflows.
Conduct low-level performance profiling analysis to identify bottlenecks in existing machine learning pipelines and propose actionable improvements.
Contribute to the development of best practices and tooling for model optimization and deployment.
Collaborate with cross-functional teams, including data scientists and software engineers, to integrate and deploy optimized models into production environments.
Partner across organizations to build tooling, interfaces, and visualizations that make the ML@Roblox a delight to use.
Qualifications
Minimum
6+ years of professional experience and a tool chest of system design experience upon which to draw to build performant systems for all of Roblox.
Have significant experience debugging GPUs - reading GPU profiles, debugging Xid errors, etc.
Proficient in advanced tools and frameworks (e.g., CUDA, Triton, TensorRT) to enhance model execution speed and reduce latency.
Experience with model optimization techniques for LLMs, such as speculative decoding, continuous batching, quantization, etc.
A Bachelor's degree in Computer Science, Computer Engineering, Data Science, or a similar technical field.
Preferred
No preferred qualifications listed.