Senior Radar Perception Engineer, Obstacle Foundation Models - Autonomous Vehicles

NVIDIA; United States; On-site; USD 180,000 - 240,000; Full time

Job summary

NVIDIA is seeking a Senior Radar Perception Engineer to architect and productize the radar-based 3D obstacle perception stack for autonomous driving. You will work on end-to-end perception with radar, camera, and lidar inputs, driving model development, data strategies, and deployment readiness.

You will lead research and production-grade implementation, focusing on scalable transformer-based models and efficient fusion techniques to improve accuracy, latency, and robustness in real-world

Qualifications

• 12+ years of hands-on experience developing deep learning-based perception, radar signal processing, or closely related systems.
• Strong proficiency in PyTorch and a track record of moving models from prototype to production.
• BS/MS/PhD in Computer Science, Electrical Engineering, Robotics, or related field (or equivalent experience).

Responsibilities

• Architect and roadmap radar-based 3D obstacle perception for end-to-end autonomous driving functionalities using state-of-the-art DNN and transformer-based architectures.
• Conduct applied research to maximize information content of radar data and tackle issues like low angular resolution and multipath reflections.
• Design and implement 3D perception models using radar inputs and multi-sensor fusion (camera, radar, lidar) for detection, tracking, and BEV understanding.

Skills

Deep learning perception

Radar signal processing

Python/C++ software engineering

Education

BS/MS/PhD in CS/EE/Robotics

Tools

PyTorch

Python

C++

Job description

Intelligent machines powered by artificial intelligence-computers that can learn, reason, and interact with people-are transforming every industry. GPU-accelerated deep learning provides the foundation for machines to perceive, reason, and solve complex problems. NVIDIA GPUs run deep learning algorithms that simulate aspects of human intelligence, acting as the brain of computers, robots, and self-driving cars that can perceive and understand the world.

We are seeking an exceptional Senior Radar Perception Engineer to help design and productize NVIDIA's next-generation autonomous driving perception stack. You will work on the core 3D radar and multi-modal obstacle perception pipeline, contribute to architecture and algorithm design, and remain deeply hands-on with implementation, including modern transformer-based, radar-centric foundation models, and multi-sensor fusion techniques where they add real value.

What you'll be doing:

• Architecture & Roadmap: Develop and improve the technical design, architecture, and roadmap for radar-based 3D obstacle perception to support end-to-end autonomous driving functionalities, leveraging state-of-the-art DNN and transformer-based architectures.
• Radar Perception Innovation: Conduct applied research on deep learning models to maximize the information content of radar point cloud data at every representation level. Tackle radar perception's hardest problems: low and non-uniform angular resolution, multipath and ghost targets, micro-doppler signatures for small targets, and severe class imbalance. Explore weakly-supervised pretraining and improve radar perception via large auto-labeled datasets.
• Model Design & Fusion: Design and implement advanced 3D perception models utilizing radar inputs (ranging from low-level range-doppler/azimuth-elevation maps to sparse/dense point clouds) and multi-sensor fusion (camera, radar, lidar) for obstacle detection, tracking, and Bird's-Eye-View (BEV) scene understanding.
• Sensor & Stack Integration: Drive radar sensor evaluation, selection, and layout optimization to support L2-L4 autonomous driving applications, ensuring seamless multi-sensor fusion.
• Production Deep Learning: Build efficient, production-grade deep learning models: define objectives with the team, select and prototype architectures, run experiments, and follow best practices for training and evaluation, using techniques such as large-scale radar pretraining, cross-modal distillation (e.g., lidar-to-radar), and parameter-efficient fine-tuning (e.g., LoRA).
• KPIs & Error Analysis: Help define and maintain KPI frameworks to quantify radar perception performance; analyze large-scale real and synthetic datasets to identify failure modes unique to radar (e.g., multipath reflections, clutter, ghost objects) and systematically improve accuracy, robustness, and efficiency.
• Data Strategy & Auto-Labeling: Contribute to the data strategy for radar perception: specify data and labeling requirements, help prioritize data collection and annotation, and collaborate with data and ground-truth teams, incorporating model-assisted workflows (e.g., active learning, automated radar labeling via lidar/camera foundation models) and model-in-the-loop tooling.
• Cross-Functional Productization: Collaborate with safety, systems, and software teams to ensure radar perception solutions meet product requirements for safety, low latency, resource usage, and software robustness, and are ready for deployment at scale.

What we need to see:

• Industry Experience: 12+ years of hands-on experience developing deep learning-based perception, radar signal processing, or closely related systems for complex real-world problems, with strong proficiency in frameworks such as PyTorch and a track record of taking models from prototype to production.
• Data-Driven Workflows: Proven experience in data-driven development, including close collaboration with data, labeling, and ground-truth teams on radar data strategy, labeling quality, and iterative model improvement.
• Software Engineering: Strong programming skills in Python and/or C++, with experience building reliable, high-performance, production-quality software.
• Collaboration: Excellent communication and collaboration skills, with the ability to work effectively across multidisciplinary teams spanning AI, hardware, and safety engineering.
• Education: BS/MS/PhD in Computer Science, Electrical Engineering, Robotics, or related fields (or equivalent experience).

Ways to stand out from the crowd:

• Radar & Multi-Modal Scale: Experience designing and deploying radar-based or multi-modal perception solutions for autonomous driving or robotics using deep learning at scale.
• Embedded Optimization: Hands-on experience architecting and deploying DNN-based perception pipelines on embedded or real-time platforms, including optimization for latency, memory, and compute constraints, and familiarity with modern architectures (e.g., Transformers, BEV networks).
• Signal Processing Depth: De