Hi Oliver, you probably have huge amounts of raw data and annotated datasets to train your algorithms. 2 questions: 1) How do you decide what you annotate and

Hi Oliver, you probably have huge amounts of raw data and annotated datasets to train your algorithms. 2 questions: 1) How do you decide what you annotate and what not? 2) How do you manage your datasets infrastructure-wise (where and how is the data hosted, what software do you use to access and filter it, ...)