Myne applies machine learning (PyTorch, TensorFlow, Keras) to automate metal recovery from recycled materials and industrial scrap, positioning itself as a non-mining source of copper, aluminum, zinc, and other metals. The tech stack reflects an engineering-first operation; hiring is sparse and concentrated in ML/software roles, suggesting the core product is still in scaling phase—projects show active work on pilot-to-production transition, process optimization, and commercial dashboards, all consistent with a company moving from lab prototype toward full-scale operations.
Myne is a Dutch metals company founded in 1917, headquartered in Harderwijk, operating at 51–200 employees. The business recovers metals from recycled and scrap sources using AI-driven sorting and processing technology, avoiding traditional mining. Product focus spans copper, aluminum, lead, zinc, brass, and bronze recovery. Current operations center on scaling manufacturing (xorter conversion system, process design, pilot testing), building commercial viability (pricing models, dashboards, loyalty programs), and expanding market reach (digital channels, international marketing). The company is a public entity.
PyTorch, TensorFlow, and Keras for machine learning models, paired with Docker and MLflow for deployment and workflow management. Label Studio is used for data annotation and model training.
Scaling the xorter conversion system from pilot to full production, optimizing processes for commercial efficiency, launching digital channels and loyalty programs, and building pricing models and commercial dashboards to improve market performance.
Other companies in the same industry, closest in size