I've worked with two popular voice recognition engines for the compact
framework. One had performance issues, the other, a leading package, does
not offer a key feature.

Thanks