With the shift in market demand towards intelligent call routing for call centres, the requirement for natural language applications is ever increasing.
"How may I help you?" is the type of open question that your caller could be faced with. Using the teliSpeech Call Router in your new services allows callers to speak spontaneously, as if they were talking to another person. According to their response, the call is routed directly to the appropriate agent or self-service application.
The teliSpeech Call Router, part of telisma's Eclipse based suite of Studio tools, opens the door to improved customer satisfaction and improved agent productivity; it prevents from customer frustration with IVR wait times,from extended call durations and from unnecessary repetition of their requests to numerous agents.
Faster and cheaper
With the teliSpeech Call Router, telisma makes it faster and cheaper to deploy Natural Language applications:
When building and deploying the application, the development phase has been cut by up to 35%: the reduction is especially dramatic in the time needed to produce the Statistical Semantic Models (SSM) and Statistical Language Models (SLM), used by the teliSpeech Recognizer to output the textual representations of any spoken input and provide the identified routing destination of the call.
The time spent during the updating phase has been reduced by upwards of 50% thanks to lower Total Cost of Ownership, reduced operating costs and a major reduction on Professional Services costs.
The teliSpeech Call Router is ideally used with the largest text corpus possible so as to best represent the typical interactions between a customer and a call centre agent.
The Eclipse plug-in offers editing and storage facilities for the entire modeling classification process. Specific editors allow you to configure the corpus analysis and modeling tasks to be carried out extremely quickly, and memorise these tasks within the project, ensuring the traceability of the models.
The Eclipse plug-in frees the user from having to know any sets of command line instructions; from now on running a training task for a model will consist of simply filling out the different parts of forms on a GUI .
Dedicated training and PS
Capitalizing on our experience we have developed specific trainings and methodology to help you through your natural language projects. We can help you answer your queries: "How large should my corpus be"? "What vocabulary and classes should I select", "How can I handle any ambiguities?" etc. telisma helps you to secure your project success.