About

Frederik Kaster is a highly skilled researcher and developer with over 13 years of relevant experience in the field of machine learning and data analysis. Currently, he holds the position of Senior Deep Learning Inference Software Engineer at NVIDIA, where he has made significant contributions to the development of TensorRT software. He has added support for dynamic shapes and optimization profiles, prepared major parts of the code base for automotive safety certification/self-driving cars, and extended the support for pointwise operations in TensorRT by over 65%. Before joining NVIDIA, Frederik worked as a Quantitative Portfolio Analyst & Software Developer at Oxford Asset Management LLP and as a Research Scientist at Mirada Medical. At Mirada Medical, he focused on medical image registration, multimodal image fusion, oncology imaging, and nuclear medicine. He specified and developed algorithms and software prototypes for computer-assisted medical diagnostics, assumed the role of product manager for Mirada's radiotherapy treatment planning software, and represented the company in the scientific and clinical community. Frederik's educational background includes a PhD in Physics from Ruprecht-Karls-Universität Heidelberg and an MSc in Physics from Johannes Gutenberg-Universität Mainz. He has a strong tech stack that includes deep learning, software engineering, research science, machine learning, testing, image segmentation, and product management. Frederik is a highly motivated individual who is passionate about using his skills to make a positive impact on the world.