This could be especially useful for any project where you don't really know much about your data set and you're just getting the lay of the land. Makes me want to throw some data from http://www.kaggle.com/ in there.
Yes, the goal is for Divvy to be a step in your workflow where you get a feel for your data. We make it really easy to export the clusterings, embeddings and visualizations you come up with to csv and png for additional analysis with other tools.