Hi everyone, I met an ML friend in NYC who has a very cool story of going from teaching himself ML to working on deep learning professionally. He’s also watched all of the fast.ai videos and is a fan 
I think his whole blog is interesting (e.g. RNN gradients from scratch, embedding airports for flight delay prediction), but this post goes over his year of doing a self-styled “Open-Source Machine Learning Masters”: http://willwolf.io/2018/02/03/practical-guide-open-source-ml-masters/. Some good advice in there about the importance of blogging and, as Rachel says, making peace with personal branding.
For those of you looking to do something similar, I hope his blog gives you some tips!