Senior 3D Perception Engineer — Foundation Models (Equity)

NVIDIA; United States; On-site; USD 184,000 - 357,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA is seeking a Senior Perception Engineer to advance its autonomous driving perception stack. You will lead the 3D obstacle perception pipeline, combining multi-camera and multi-sensor fusion to deliver robust object detection and tracking in real time.

You will craft production-grade DL models, optimize for latency on GPUs, and collaborate across data, safety, and software teams to ensure scalable deployment. Strong experience with PyTorch and CUDA is essential.

Qualifications

• PhD/MS/BS with extensive experience in computer science or related field.
• Hands-on experience with deep learning perception systems and production-quality software.
• Strong Python and/or C++ programming skills.
• Experience with PyTorch and GPU-accelerated training/inference.

Responsibilities

• Design and implement 3D obstacle perception models using multi-camera or multi-sensor fusion.
• Develop production-grade perception pipelines with state-of-the-art CNNs and transformers.
• Define KPI frameworks and analyze large datasets to improve accuracy and efficiency.
• Collaborate with safety, systems, and software teams to meet product requirements.
• Contribute to data strategy, labeling, and model-assisted workflows.
• Push for BEV and transformer-based 3D perception in real-time systems.

Skills

Python

C++

PyTorch

Communication

Education

PhD

MS

BS

Tools

CUDA

Job description

NVIDIA is seeking a Senior Perception Engineer to advance its autonomous driving perception stack. You will lead the 3D obstacle perception pipeline, combining multi-camera and multi-sensor fusion to deliver robust object detection and tracking in real time.

You will craft production-grade DL models, optimize for latency on GPUs, and collaborate across data, safety, and software teams to ensure scalable deployment. Strong experience with PyTorch and CUDA is essential.