Machine Learning Engineer, II - 3D Perception

Job description

Job responsibilities

• Design, develop, and improve machine learning models supporting Torc's perception systems.
• Own model development and delivery for well-defined perception problem areas, from data preparation and training through evaluation and integration.
• Write production-quality Python and PyTorch code to support scalable training, evaluation, and inference workflows.
• Analyze model performance, identify failure modes, and independently troubleshoot issues to improve robustness, accuracy, and generalization.
• Develop and evaluate perception models leveraging multi-modal sensor data, with an emphasis on camera-based and 3D perception systems.
• Collaborate with software engineers, infrastructure teams, and autonomy engineers to integrate perception models into larger production software systems.
• Contribute to improvements in training pipelines, data workflows, experimentation tooling, and developer workflows that accelerate model iteration and deployment.
• Participate in model architecture discussions and contribute technical recommendations within the team.

Job requirements

• Bachelor's degree in Computer Science, Robotics, Electrical Engineering, Machine Learning, or a related technical field with 4 + years of relevant industry experience, OR Master's degree with 3 + years of relevant…
• Experience developing machine learning models for computer vision, perception, robotics, autonomous systems, or a closely related domain.
• Strong programming skills in Python and PyTorch, with experience writing maintainable, production-quality machine learning code.
• Experience training, evaluating, and improving deep learning models using large-scale datasets.
• Experience working with image-based and/or 3D perception systems.
• Solid understanding of deep learning architectures commonly used for perception applications.
• Experience debugging model behavior, analyzing performance metrics, and proposing practical improvements.
• Ability to independently execute complex machine learning work within well-defined problem areas.