Recruiters love side projects because it shows initiative. At least for data science (which I've been considering moving into after I finish my dissertation), they like if you have projects that are available on the web--code on github, writeup on a blog, and/or posts on twitter. Plus it's more fun and easier than taking structured classes since you can do the projects on whatever interests you. I've also considered a data science bootcamp but decided that it isn't for me right now. Another thing that people want is some experience with machine learning--it seems that many people interested in the field started by taking one of the MOOCs, usually the Andrew Ng Stanford one although I recently talked with a data scientist who preferred the Udacity one.