I think in this particular case, the problem was just picking the wrong tool for the job. Python is the primary language used for data science these days, but it is not particularly designed for it. It is a general purpose programming language. If they had chosen something like R, their path would have been a lot smoother.