From what I hear this is a pretty powerful data mining package (developed at the faculty I attend). Personally I haven't gotten to play around too much with it, never had the time to figure it out properly.
Protip: use it as a python library, a lot of stuff doesn't make it into the GUI.
It looks really cool; is it going to be the WEKA killer for '10?
Speaking as a rubyist, now that's a compelling reason to use Python more.
I would be happy to use this to read some MongoDB data as well (I guess it's probably possible).