1. How to make the most out of your undergraduate degree?. + NETWORKING!
2. If you are a Machine Learner, why should you consider learning atleast some systems knowledge? Read this blog. Also stop using Windows.

3. If you're interested in and want to learn Reinforcement Learning, consider taking a look at GenRL. It is a Reinforcement Learning library that is easy to understand i.e., accessible and made for people who want to get into Reinforcement Learning Research. We have big plans. :)

4. Since we're talking about Systems and Reinforcement Learning, have you taken a look at QuaRL? Faster decisions, makes everyone happy!

5. If you're interested in terra scale Machine Learning (too many numerical features, etc., magnitude order >> 3), you should check out VowpalWabbit. It is also the most optimized (for performance) Machine Learning library that I've ever seen. Hosts major Contextual Bandit algorithms because of their common founder - John Langford.

6. To this day, probably the most decorated success story of the application of Modern Machine Learning Reinforcement Learning in the industry (aka real world) is probably Contextual Bandits applied to advertisement, recommendation, etc. There are a lot of reasons to that. Few of these qualities include sub-linear (logarithmic) learning (regret), extremely fast decisions, ability to scale and deploy easily, etc., where most ML algorithms suffer.