At Universal Robots I was employed in the Research and Development department. Here I worked with C/C++ on kernel-level and embedded robotics systems, creating crucial insight and domain knowledge which could be leveraged to support feature engineering for our predictive models.I further developed user-facing applications for our robotics controller which facilitated our data-collection process and in addition to this, utilized my knowledge of the data science domain to develop predictive maintenance models for robotics. These models - implemented in Python using technologies such as TensorFlow and sklearn - were designed to mitigate the costs from downtime in operations, by allowing us to send out technicians before the robots fully broke down.

At Findwise, I played an active role in product ideation, and successfully implemented several state-of-the-art deep learning and NLP techniques to bring these products to life. This involved the development,parameterization and training of neural network models - typically for temporal data such as natural language. Additionally, I designed Computer Vision algorithms with the purpose of extracting relevant metadata from complex file types such as scanned documents and blueprints.In conjunction with my technical work, I also provided on-site consultancy services to our clients. Doingon-premise development, taking part in meetings and helped carve out the road-map of our future data-science product chain.