Vestec logo
India USA
China Australia
 
main splash image
Robust Affordable Speech Recognition

   • $199 per channel perpetual license
   • Acoustic models for 4 major languages:
   - American English - Chinese Mandarin
   - Australian English - Indian English
   • SRGS (XML & ABNF) grammar format
   • MRCP (v1 & v2) platform integration
   • Scalable architecture & load balancing
   • Asterisk and FreeSWITCH support
   • Linux and Windows installation

KE Benefits


The KE System enables faster time-to-market for natural language call-steering applications and reduces grammar development and maintenance costs compared to conventional approaches. In addition, it significantly improves caller experience relative to traditional DTMF menus and discrete-speech, directed-dialog systems.

Grammar development time and costs are reduced by: (a) lack of need of tagged transcriptions data for training; (b) no requirement to write semantic parser for interpreting recognition results; (c) powerful graphical interface for designing call-flow; (d) transparent keywords-based grammar for routing vocabulary specification; and (e) robust speaker and domain independent garbage model for detecting specified keywords in caller utterances.

Caller experience is enhanced relative to DTMF menus via customer ability to: (a) speak in natural conversational language; and (b) gain direct access to relevant menu option. Compared to conventional discrete-speech implementations, the KE system allows callers: (a) to use specified keywords in utterances of their choice; (b) avoid tedious level-by-level menu traversal; and (c) replace directed-dialogue interface with an open-prompt.

 
 
 
©2012 Vestec, Inc. All rights reserved.