VoiceXML Quick Start Guide

by Jonathan Eisenzopf

If you want to develop VoiceXML applications but don't know where to start, this guide is for you. This 8-step guide contains all the information you need to start developing and deploying VoiceXML applications—prior experience is not required.

If you want to develop VoiceXML applications but don't know where to start, this guide is for you. This 8-step guide contains all the information you need to start developing and deploying VoiceXML applications--prior experience is not required.


You will need the ability to transfer files via FTP to a Web server that is accessible on the Internet. You will also need a computer running Microsoft Windows NT, 2000, or XP with at least 128MB of RAM and at least 200 megabytes of free disk space. A Pentium II 350MHz processor is the bare minimum amount of CPU power that you will need. You will also need a Sound Blaster compatible sound card.

Step 1 - Download and install Nuance V-Builder

Open a Web browser and go to http://extranet.nuance.com. If you haven't registered as a member of the Nuance Developer Network, click on the link that is labeled "Sign Up", otherwise login to the extranet site. Once you've logged in, click on the Download link, then click on the Nuance V-Builder link. Download the file named V-Builder-1-2-GA.exe. The program is about 25 megabytes, so it'll take a while to download if you're not blessed with a fast Internet connection.

Once you've completed the initial installation, launch V-Builder and select Update Center in the Tools menu. You will be prompted to enter your Nuance developer login and password. V-Builder will connect to the V-Builder update server and return a list of modules, updates, and plug-ins that are available. Before you can run any VoiceXML applications, you will need to download a language pack. Language packs enable the speech recognition functionality in V-Builder. For example, I downloaded the English.America Master Package to be able to recognize American English speech. You will also need to download the Nuance Java SpeechChannel Pack and the Nuance Core Recognizer Essential Pack

To enable text-to-speech functionality from within V-Builder, you must manually download VOCALIZER-1-0-SP3-NT.zip from the Nuance Developer Network Web site. Vocalizer includes an American English synthesizer, but you will need to download a language pack from the Vocalizer download site for other languages. Unzip the file, run Setup.exe, and follow the instructions. The package will also install a copy of the Java 2 run time environment.

You now have all the tools you need to develop a VoiceXML application. Now move on to Step 2.

Step 2 - Create a V-Builder project

 The first time you start V-Builder, you will see a welcome screen.


Select the first link to create a new project. You will be prompted to create a new project directory that will hold the project dialogs, grammars and audio files.

Click on the Create button to accept the default directory. You can, of course, choose a different project directory and name if you like.

Step 3 - Modify the main.vxml dialog

Once you've created a new project, double-click the main.vxml dialog in the project tab.

V-Builder includes the ability to drag and drop VoiceXML elements without having to know the language.


The main.vxml dialog includes a default form. You will need to make the following modifications:

  • Re-name the form to "main" by selecting the properties tab and changing the name attribute from "form1" to "main." 

  • Drag and drop a field object into the form.

  • Drop a grammar, prompt and filled object into the field object.

  • Drop a text object into the first prompt object and enter the following phrase in the text area of the general tab on the right, "What is your favorite color? Red, green or blue?" This will prompt the user to say their favorite color. 

  • Drop a prompt object into the filled object. This prompt will play back the color that the user selected. 

  • Drag a text object into the prompt object and enter "You said" in the text property text area. 

  • Drop a value object into the prompt object after the text object and set the expr attribute to =color in the property tab. This will output the user's selection that has been stored in the color field. 

  • Drop a clear object into the prompt object after the value object and set the namelist attribute to include the field color. This clears out the color field. 

  • Finally, drop a goto object into the prompt object after the clear object. Click on the "Select dialog and form" option and select main.vxml from the dropdown.


Now it's time to fill in the grammar for the colors that the system will recognize. For this example, we'll limit it to red, green and blue.

  • Select the grammar object and click on the "Use Inline Grammar" option.

  • Click the "Edit Inline Grammar" button. 

  • Right click on the grammar window and select "Insert New Expression". 

  • Enter the words red, green and blue in separate "Phrase Text" fields and select the OK button. To add more words, click the "Add Another Row" button.


The grammar diagram should now appear as below. This grammar will match any of the three words in the list.

Your first VoiceXML dialog is now ready for testing. Save the file and the proceed to Step 4.

Step 4 - Test the dialog

Now that the dialog has been completed, it's time to test it within V-Builder by pressing the button to run to dialog. If this is the first time the dialog has been tested, you will have to wait a few minutes for the speech recognition and TTS servers to start up. Once the servers are ready, you will see a green light in the "Run Dialog" window. Click the "Run Dialog" button. You will hear a computerized voice prompting you to select red, green, or blue. When you hear this, speak one of the colors into the microphone on your computer. The Nuance recognizer will fill in the color field with the recognized value, play it back and prompt again for a color until you click the "Hang Up" button. 

Continue on to Step 5 or repeat Step 3 and Step 4 until you are satisfied with your new VoiceXML dialog. 

Step 5 - Upload the dialog to a public Web server

Once you're satisfied with the voice dialog, you will need to upload the main.vxml file to a Web server that is publicly available on the Internet. Make sure you know the full Web URL to main.vxml on your Web server after it has been uploaded. If you've uploaded the file into your Web root directory, and the hostname of the Web server was www.myhost.com, then the URL would be http://www.myhost.com/main.vxml. Once you've uploaded the file and have verified that you have the correct URL to the file, proceed on to Step 6.

Step 6 - Set up your Voxeo developer account

  You can be assured that voice dialogs developed in V-Builder will work on the Voxeo platform because they utilize the Nuance Voice Web Server. Voxeo is a Voice ASP, which means that they manage the hardware and software and provide you with a toll-free phone number to access your VoiceXML application. Customers dial the toll-free number which connects to a Voxeo gateway. The gateway then interacts with your VoiceXML application, which resides on a Web server somewhere on the Internet. You will first need to register for a free developer account. Then you can login to your account on the Voxeo Community Web site, http://community.voxeo.com.

When you login, you will be greeted by the Voxeo Account Management screen as seen below. Click the <add new app> link.

View Voxeo Account Manager Screen Shot

Select the Nuance Voice XML v1.2 Development option and click the "NEXT" button.


Now, select a name for your application, enter the full Web URL to the main.vxml file on the Internet, and click on the "FINISH" button.


You will see a confirmation screen that includes the phone number and PIN code for your application. Write this information down for reference and click the "NEXT" button. Finally click on the button labeled "SAVE CHANGES". You will see that the application has been added to the Development menu.

Step 7 - Test the application on the telephone

If the URL you entered was valid, then you can now dial into the application on any telephone using the toll-free number and your PIN code. Repeat steps 3 through 7 until you are satisfied with your application.

Step 8 - Deploy the application

Once you're satisfied with your application, contact a Voxeo sales representative who will assign a dedicated phone number to your application and give you access to high speed servers and outbound dialing capabilities such as call transfer.


You now have the platform upon which you can begin developing VoiceXML applications. To learn more about the VoiceXML language, I would recommend reading A Tour Through VoiceXML, a 12 part series that provides a thorough introduction of the language. If you have any other questions regarding V-Builder, Voxeo, or have problems implementing the steps in this guide, let me know by sending an email to eisen@ferrumgroup.com.

About Jonathan Eisenzopf

Jonathan is a member of the Ferrum Group, LLC  which specializes in Voice Web consulting and training. He has also written articles for other online and print publications including WebReference.com and WDVL.com. Feel free to send an email to eisen@ferrumgroup.com regarding questions or comments about the VoiceXML Strategy series, or for more information about training and consulting services.

This article was originally published on Saturday Sep 28th 2002
Mobile Site | Full Site