The easiest way to understand how a VoiceXML application accesses information that is stored in back-end databases is to consider how such transactions are performed in an HTML Web page. In an HTML Web page, you present a form where the user can type in the information required by the transaction. The user then clicks on a Submit button. This invokes an HTML <post> tag, which causes the form to be posted to the Web server, where it is processed by a server-side application.
In a VoiceXML document, the form is presented to the user by prerecorded audio or text–to–speech. The user provides input and makes choices by using speech or by pressing keys. A VoiceXML document is simply an auditory version of a form. When the form is complete, the input is posted to the Web server by using a <submit> tag in the document. On the Web server, the form can be processed by the same server-side application that processes the equivalent HTML form.
In both cases, the server-side application processes the user’s input, accesses the back-end server database with an SQL command or similar, updates it or obtains the required information, and returns to the VoiceXML application, which presents the outcome to the user using prerecorded audio or text to speech.