Robust Affordable Speech Recognition
KE Overview
The KE System is a grammar generation and NLU engine for developing natural language call-steering applications without the conventional need for transcriptions-based training data and SLM grammar-based speech recognition engines. It is well suited for applications involving up to 15-20 routing destinations and works as an add-on to large vocabulary 3rd party finite-state grammar speech engines. It significantly reduces the time, effort, and cost for developing natural language grammars and provides a powerful means for speech-enabling traditional DTMF menus or for offering a sophisticated alternative to conventional discrete-speech, directed-dialog implementations.