About VoiceXML

VoiceXML is a language of the XML family. It is specifically designed to describe voice user interfaces.  Its primary goal is to bring to speech applications the methodologies and tools which made the success of the Web.  VoiceXML actually is to speech applications what HTML is to visual applications.

The main strength of VoiceXML is in standardization.  A traditional IVR application can function only on servers of the operator owning its proprietary technology.  On the contrary, VoiceXML applications are portable.  They can function on the most appropriate servers at each stage of their lifecycle.

A VoiceXML application is made of multiple resources:

• VoiceXML scripts, which define the contents and sequencing of dialogs between the user and the application.
• Pre-recorded audio, which the application plays during the dialogs.
• Text, which the application plays with speech synthesis.
• Speech recognition grammars.
• ECMAscript resources which make up VoiceXML dynamic capabilities.

Those resources are stored in an application server.  They are loaded via HTTP, HTTPS or, optionally, other protocols.  They are played by a dedicated server called "VoiceXML Gateway".  This is the only server actually connected to the telephone network.

The resources may be static of dynamic.  Interpretation of those resources by the VoiceXML Gateway lets users access the desired interactive content.