Senior Software Engineer, Cyngn

• MS/PhD in computer science, computer engineering, robotics, or similar technical field of study
• 4+ years of experience writing Python software in a production environment - unit testing, code review, algorithm performance trade-offs, etc
• Strong theoretical foundation in deep learning techniques for computer vision, with working knowledge of linear algebra, probability, and optimization
• Hands-on experience developing and deploying deep learning models for real-world perception tasks (e.g., detection, segmentation, multi-object tracking)
• Proficiency with libraries such as Pytorch, TensorFlow, Numpy, SciPy, OpenCV (Python), etc
• Experience building and integrating tools and infrastructure to optimize model development lifecycle, including but not limited to model versioning, model evaluation, model deployment, etc
• Excellent written & verbal communication skills
• Additional experience with model acceleration techniques, building model development infrastructure, and ownership of the end-to-end model lifecycle—from research and training through deployment in production—is considered a strong bonus

Desirable

• We are seeking an experienced Senior Software Engineer to join our Perception team within the Autonomy Software organization
• This role centers on developing advanced deep learning models that power object detection, segmentation, and tracking using camera and lidar data for Cyngn’s autonomous forklifts
• A key responsibility will be designing, training, and deploying multi-modal perception models to detect and localize pallets in diverse warehouse environments, enabling reliable pallet engagement and stacking
• Design, implement, and optimize deep learning models for object detection, segmentation, and tracking using camera and lidar data
• Build and maintain data pipelines, training infrastructure, and inference frameworks to support reproducible and scalable model development
• Develop tools and metrics for evaluating model performance and ensuring robustness across diverse warehouse environments
• Work with third-party annotation vendors to generate high-quality labeled datasets for training and validation