Congratulations Dan on the release. This does fill a gap, both in terms of developers' skills and existing toolsets' availability and functionality. My question would be, could this be used out of the box in educational environments, for instance to assess accuracy of pronunciation?