I'm thinking ahead for the company I work for. I was wondering how we could organize and store datasets for the long run. Because, yeah, we're a bit late on dat

I'm thinking ahead for the company I work for. I was wondering how we could organize and store datasets for the long run. Because, yeah, we're a bit late on data management. So I work in a Machine Learning department and getting our hands on data from clients have been a terrible mess (both architecturally and legally). How are we suppose to develop prediction machines for contexts we have no data on? I wish to push the discussions a bit in the right direction.