Wind Telecom - Contact Center, Call Center, IVR

Voice services platform

When we make a phone call to a modern company we are most often welcomed by a pleasing voice encouraging us to stay in the service and giving us instructions how to get to a desired information or contact an agent.

Until we select an option to contact the agent who will welcome us and answer our questions, it is the automatic machine that is responsible for our communication with the service. This automatic machine holds conversation with a caller through presenting information available in the form of sound files and providing options available on the consecutive menu levels. The machine is listening to the options that the caller selects by pressing particular keys on the telephone keypad through which he or she indicates what they are interested in and what’s the purpose of their call. Until recently, it was a typical scenario of customer care service.

This mysterious automatic machine operating strictly in accordance with the programmed algorithm through which we can access specific functions on the defined menu levels is called interactive voice response (IVR) system.

A question arises whether IVR systems supplied with the recorded voice announcements in the form of sound files or recognition of DTMF codes of the pressed keys have set the limits for the described technology. Or maybe there are some other possibilities?

Technology has gone much further. Nowadays voice services using text-to-speech (TTS) technology in order to read out information stored in the form of text files are standard services. Moreover, another technology called automatic speech recognition (ASR), which we only mention in this article, and which allows speech recognition and holding conversation with a caller without pressing any key, is currently taking over the market.

IVR services supported with TTS technology are more and more frequently applied in contact center solutions, where they relieve agents of answering particular types of calls through providing the desired information automatically via voice services or specialized services, thus shortening the time of customer care service and reducing the costs.
Currently about 40% of IVR systems use text-to-speech technology, which affects both functionality and quality of the voice systems.

The speech synthesizer technology has been known for a long time, its quality evolves with time and more and more frequently it becomes difficult to recognize whether the text of a welcoming message comes from a tape or it is dynamically generated by the system. This is due to the fact that algorithms of generating text-based sound files are better and better, and the competition on the market of advanced voice technologies is growing.

Interest in this technology is driven by many factors, the most important of which are the very high quality of the generated voice and eluding similarity to the human voice.

Moreover, another factor interesting to the customers is the dynamic presentation of messages stored in the form of text in database, file on the disk, or announcement on the website. It is worth underlying that voice service users can take advantage not only of a general information available to any caller, but also of information dedicated to a particular person. Realization of this functionality on the basis of the standard technology based on pre-recording of messages is practically impossible due to the large scale of the undertaking.
In case of application of TTS technology information dedicated to a person verified in the voice service on the basis of the phone number or PIN code entered via phone keypad can be generated dynamically.

High quality of TTS technology does not only stand for a nice and fluent voice of a synthetic consultant that sounds like human, but also correct reading of dates, time, abbreviations and other elements indispensable for language definition. Openness to new foreign language system versions and to various sounds of male and female voices is of equal importance.

Top page