For clarity, I only covered the data science part of the perspective; the data engineering/DevOps part is another, more difficult can of worms which would require its own post!