Public summary
Join a health technology company developing digital and robotic medical devices for musculoskeletal rehabilitation. As a Senior Machine Learning Engineer, you will build and deploy edge AI models that evaluate patient exercise quality using computer vision and deep learning. You will collaborate with physiotherapists and doctors to translate clinical evaluation criteria into AI-driven feedback. Work hands-on on model training, optimization, deployment on mobile devices, and support body and hand tracking model improvements within a dynamic, cross-functional team based in Mainz, Germany.
Location and work setup
- Location
- Mainz, Wiesbaden
- Remote status
- Remote
- German requirement signal
- No German Required Detected
- Detected job language
- English
Responsibilities
Design, implement, and deploy deep learning models for evaluating rehabilitation exercise quality from video keypoints. Handle large-scale, multi-class fault detection generalized across many exercises. Address challenges such as noisy data, occlusions, and low-latency requirements for edge devices running Android and iOS. Own end-to-end data pipelines including data extraction, annotation strategy, model development, testing, and production deployment. Research novel approaches to improve model generalization and efficiency. Collaborate closely with clinical experts to align AI models with medical standards. Support and optimize existing body and hand tracking models alongside the data science team.
Qualifications
Proven experience building and deploying deep learning and computer vision models with end-to-end ownership. Expertise in Python with PyTorch or TensorFlow. Experience with time-series or video-based action quality assessment, e.g., monitoring assembly steps, surgical skill, or ergonomics from video. Practical knowledge of transfer learning and few-shot learning to generalize model predictions to new motion or fault classes. Experience in model compression techniques including quantization and pruning for deployment on resource-constrained edge devices. Bonus: Background in biomechanics or physiotherapy, experience training pose estimation or hand tracking models, and familiarity with medical or regulated environments.