End-to-end data science projects encompass one or several full iterations of the data science project life cycle. End-to-end data science projects comprise of all the risks of research projects and MVPs, along with a new set of risks related to change management and production deployment.
The first major risk is the inability to sustain a constant change stream. Data science projects involve scope changes, and you should be able to work with them without making the project fall apart. Scrum gives you the basic tools you need for change management by freezing the development scope over the course of the week. However, for any tool to work, your customer should understand and follow the required processes, along with your team.
Another issue is that the implementation of a given change may cause a lot of unexpected bugs...
