In 2008 when I was on the Bing team, Sergei had just developed this library and was putting it to work on the document understanding pipeline. He was incredibly passionate about it and would evangelize it with anybody that cared to listen. When I saw the headline on HN, I immediately guessed that this was Sergei's work. And of course it is! Sergei, if you are reading this, try to guess who I am ;)