"Hey Facebook" speech commands help with streaming chunked 10 sec http REST GET in my game

Just did the F8 hackathon, my intermediate at best knowledge of C# and many internet searches failed me in one respect. Try as I might have I could not save a wav with any voice data. I used some code from lipsync and got sound, could create a wav file header but no sound data.

Tried many methods. So, instead I'd like to figure out how to stream the data to wit.ai.

If you have any idea how to make this happen? Hackathon done but I need some help making this work.

How does FB make it work? Also I'd like to add automatic activation in my app of this feature without having to click anything like it is available globally and my wake message can be activated.