What they said.
Sorry for the confusion. To sum up a little differently, there are three pieces involved:
1. Get the data IN to the VTT (which most already provide, but in very different ways)
2. Get the data OUT of HLO
3. Translate the data that comes out of HLO into the format needed for input by the VTT
We're focused on that second step right now. The third step will be different for every VTT. And we'll be looking to the VTT users who know how to do the first step to assist with the third one.