Qualifications

• 5-7 years of experience in software development, particularly in Python.
• Practical experience in object detection or instance segmentation with rigorous metric evaluation.
• Proven MLOps practices and deployment experience with PyTorch, ONNX, TensorRT, or equivalents.
• Familiarity with domain shift, imbalanced data, or active learning techniques.
• Strong software engineering skills, including modular design and automated testing.

Responsibilities

• Train, evaluate, compare, and deliver detection and segmentation models like YOLO and Mask R-CNN; diagnose errors and report detailed metrics.
• Manage real-time inference at the edge, conduct parity testing, optimize latency and throughput for deployments.
• Develop robust evaluation protocols to ensure generalization across varying conveyor qualities, lighting, and camera conditions.
• Prioritize informative images for active annotation and verify pre-annotations quality from the models.
• Specify and validate RGB-D cameras and optics under operational conditions.

Benefits

• Comprehensive oversight of robotic picking performance from sensor to production robot.
• Direct impact from technology deployed across multiple countries to support the circular economy.
• Four weeks of vacation plus a leave day on your birthday, comprehensive insurance plans, and team-building activities.