> If it's truly open, including the underlying recommendation models and algorithms, I will be happy to use it and set up a monthly donation for whoever wants to keep working on it.

Knowing examples such as Hudson CI & co, that probably makes it "no one", at a statistical scale.