Senior Research Engineer, Interactive World Models
Core
Lead engineering for NVIDIA's real-time world simulation stack (FlashDreams/FastGen), enabling autonomous driving, robotics, and virtual training applications.
Role type
Senior Research Engineer (Applied ML & Systems)
Builds
Real-time world model execution layers, continuous autoregressive serving loops, and simulation workflows.
Domain
Generative AI, World Models, Robotics, Autonomous Driving
Deliverable
production ML models | product features
Required skills
Video/World models, Diffusion/Generative modeling, Model distillation, Simulation, Robotics, Computer vision, Real-time stateful ML systems
Preferred skills
Post-training generative video models, Real-time stateful generative inference systems, Open-source ML project stewardship
Technologies
Python, PyTorch, GPU-accelerated training/inference, vLLM, SGLang, FastVideo, LightX2V, Diffusers, FlashInfer
Responsibilities
Build and optimize continuous autoregressive serving loops with per-step control and KV-cache management; Advance production-ready world models via few-step distillation and long-horizon spatiotemporal memory; Lead end-to-end delivery of multi-user simulation workflows from prototype to release.
Seniority
Senior, hands-on IC
