Senior Perception Engineer

We are looking for a Senior Perception Engineer to train and optimize the deep learning models that let our autonomous platforms see — across urban, closed community, and off-road environments. We are looking for a Senior Perception Engineer to train and optimize the deep learning models that let our autonomous platforms see — across urban, closed community, and off-road environments.

Role Responsibilities

• Train and optimize deep learning models that process multi-modal sensor data, including LiDAR point clouds, multi-view camera images, and radar signals.
• Work on multiple autonomous driving stacks including urban, closed community, and off-road environments.
• Develop and improve key perception tasks, including 3D object detection, 3D semantic segmentation, multi-object tracking, parking slot detection, and cost map.
• Work with large-scale datasets and state-of-the-art machine learning models using in-house compute infrastructure.
• Validate, benchmark, and iterate on perception algorithms through real-world testing on self-driving vehicle platforms.

Required Qualifications

• Minimum 3+ years of experience in perception systems.
• Bachelor’s or Master’s degree in Computer Science, Robotics, Mechatronics, or related field.
• Hands-on experience with deep learning applied to at least one sensing modality (camera, LiDAR, or radar).
• Familiarity with modern deep learning architectures, preferably including transformer-based models (e.g., Vision Transformers – ViT).
• Strong practical experience with PyTorch.
• Proficient programming skills in Python.
• Strong teamwork and communication skills, with a proactive and learning-oriented mindset.

Preferred Qualifications

• Strong programming skills in modern C++ 14/17.
• Experience working on autonomous driving or robotics perception systems.
• Experience with optimized model deployment frameworks such as TensorRT, ONNX, or LibTorch.
• Proven experience building and maintaining scalable ML systems using collaborative practices such as MLOps, data/version management, and CI/CD pipelines.

#J-18808-Ljbffr