As a Senior Applied Research Engineer on NVIDIA’s cutting-edge team, you will focus on advancing world foundation models for high-fidelity video generation with an emphasis on human-centric quality. Your daily tasks include researching, implementing, and validating model improvements in areas such as spatial multimodal modeling, modality alignment, and neural rendering-inspired representations to enhance controllability and long-horizon consistency. You will also optimize training and inference efficiency through architectural techniques and develop detailed benchmarks for evaluating world foundation models. The ideal candidate holds a PhD or equivalent experience, with expertise in deep learning, Python, PyTorch, C++, CUDA, and distributed training workflows. This role requires hands-on experience improving generative models for perceptual quality and temporal stability, particularly for human motion and interaction dynamics across real-world and synthetic data, contributing to the development of robust implementations that showcase capability gains across teams.
Skills
Python
PyTorch
C++
CUDA
Distributed Training
Multi-GPU
Multi-Node
Inference Optimization
Perceptual Metrics
Experiment Tracking
Deep Learning
Generative Models
Neural Rendering
Diffusion Models
Flow-Based Models
Modality Alignment
Spatial Multimodal Modeling
CI/CD
What you'll do
Research and implement model architecture changes to enhance video generation fidelity.
Prototype improvements in spatial multimodal modeling and flow-based video generation techniques.
Optimize training and inference efficiency through architectural and post-training methods.
Define training objectives for better sim-to-real generalization, focusing on human dynamics.
Develop detailed benchmarks for evaluating world foundation models' performance.
Translate research findings into robust production implementations like training code and demos.
What we're looking for
PhD in Computer Science, Graphics, or related field with 8+ years of applied research experience.
Direct experience designing, training, and evaluating generative models for image/video/audio.
Advanced proficiency in Python, PyTorch, C++, CUDA, and distributed training workflows.
Hands-on experience improving perceptual quality and temporal stability in human-centric models.
Strong track record with publications in top conferences like NeurIPS, CVPR, ICLR.
Practical knowledge of inference/runtime optimization techniques and diagnosing visual artifacts.
Market check
Salary context
This
$184,000–$287,500
range sits above
69%
of similar postings on FindRole.
Peer median band
$172,850–$257,050
Median floor and ceiling across peers.
Typical midpoint (25–75%)
$175,500–$246,150
Middle half of comparable postings.
Based on 240 comparable postings.*
* 240 is the maximum number of comparable postings sampled.
Employer
About Nvidia
Nvidia is a leading designer of graphics processing units (GPUs) and system-on-chip units, powering gaming, professional visualization, data centers, and artificial intelligence workloads. Industry: Semiconductors & AI Computing
Nvidia currently has
801 open roles
on FindRole.
Listed pay typically runs
$184,000–$287,500
across 797 roles with salary data.