I just posted an article describing my analysis of a public talk I gave previous month.
I analyzed my Fitbit heart-rate data and the transcript of my speech (obtained using Watson Speech-To-Text service).
The analysis is relatively technical, and accompanied by a Python Notebook with all the code I used to obtain such results.
I welcome all kind of comments and critiques, especially if you point out what I did wrong or could do better.
Great project, I found it novel and interesting. I especially like that you didn’t feel a need to force a dramatic lesson of some kind, and were really transparent and informative about your methods. This is valuable! Thank you. I hope you are coming to Amsterdam in June, maybe you can give a talk about your data while giving talks. 
Hi Gary,
Really appreciate your feedback.
Agree on that, I was so focused from the beginning on the pure analytical component, that I subconsciously ignored possibility for speculative lessons that are generally much looked for. Truth is, I believe Data has a much higher chance to be right than my words…
About the conference, I’ll see what I can do, and sure I admit it will be easy for me to go meta with such topic! 