Today advice

As a data scientist, you have to know that people awareness (some times including you organisation management) of the data science capabilities are not accurate. At some points they ask for very humble objectives just because they think it is the best that can be achieved, some other times they will watch a science fiction movie and ask you to do something amazing that might save the world (but it is just not possible with the data you have). Your first task when you receive a specific mission is to enlighten stake holders of what can be done -even if they do not think it cannot be done- and to tell them what is not possible and why. From my experience within EC funded projects worldwide, many projects end up with great solutions that are not exactly the best we can offer (just because we adapted to the stakeholders expectations of what can be done). Even in private countries, when we were forecasting the gas prices for example, solutions could produce exactly the right prices ± a very small fraction. But they still insisted to have general directions only, and could not believe what their own eyes see. The man-machine confidence and the stakeholders expectations will greatly impact your missions and your work. The best, is to disseminate the basic knowledge between all stake holders in your organisation (feel free to use the power of your manager, s/he will definitely help).