HeyGen · posted · checked
Software Engineer, GPU Performance
PythonPytorchOn-site · Los Angeles, Palo Alto, San Francisco, Toronto and Singapore
Who can apply
On-site · Los Angeles, Palo Alto, San Francisco, Toronto and Singapore
Location: Los Angeles, Palo Alto, San Francisco, Toronto, Singapore
About HeyGen
At HeyGen, our mission is to make visual storytelling accessible to all. Over the last decade, visual content has become the preferred method of information creation, consumption, and retention. But the ability to create such content, in particular videos, continues to be costly and challenging to scale. Our ambition is to build technology that equips more people with the power to reach, captivate, and inspire audiences.
Learn more at www.heygen.com. Visit our Mission and Culture doc here.
Position Summary
HeyGen is building AI applications including Avatar IV, Photo Avatar, Interactive Avatar, and Video Translation. We’re looking for a Software Engineer focused on GPU performance to make the systems behind these experiences faster and more efficient.
You will work across model execution and inference infrastructure, using profiling and measurement to improve latency, throughput, and GPU cost. This role is a fit for an engineer who enjoys understanding how software uses the hardware beneath it.
Key Responsibilities
Use NVIDIA Nsight Systems, Nsight Compute, and PyTorch Profiler to investigate GPU utilization, kernel execution, memory bandwidth, and CPU–GPU data movement.
Identify bottlenecks across model execution, preprocessing, and inference serving, then measure the impact of each optimization.
Improve performance through batching, scheduling, memory management, and better GPU utilization.
Develop or integrate high-performance GPU kernels when existing implementations limit performance.
Build benchmarks and automated checks that catch performance regressions across representative video workloads.
Collaborate with AI researchers and infrastructure engineers to bring optimizations into production.
Measure the effect of changes on latency, throughput, cost, and output quality.