Technical Program Manager, Model Performance
Core
Build zero-to-one planning structures and execution processes for high-performance AI inference algorithms and methods.
Role type
Zero-to-one Technical Program Manager (Model Performance)
Builds
Core algorithms and methods for high-performance inference stack
Domain
AI Infrastructure / Inference Optimization
Deliverable
production ML models
Required skills
Technical program management, Model performance/inference optimization, Zero-to-one program building, Cross-functional alignment, Risk management, Organizational design
Preferred skills
Model release coordination, Software engineering background, Deep learning performance optimization
Technologies
vLLM, TensorRT-LLM, SGLang, NVIDIA Dynamo
Responsibilities
Own execution across active project portfolio, Design planning structures and operating cadences, Coordinate model release and optimization programs, Drive cross-team alignment, Surface risks and dependencies, Partner with engineering leads on team structures
Seniority
Individual Contributor, Zero-to-one builder
