VXML breathes new life into phones

Move over, plain old phone conversations. Rajat Tandon says that phones can be now be used for everything from accessing customised information to management of a factory, all thanks to VXML

What if your telephone could give you access to a wealth of resources through your Web infrastructure? Or if callers could access all the information on your Web servers from their telephones? Or if you could share your website with people who didn’t even have Internet access?

These scenarios are no longer fictional. These scenarios are one big leap forward from the basic interactive voice response (IVR) applications, and the technology to achieve them is available right here today, in the form of Voice eXtensible Markup Language (VXML).

VXML, which is revolutionising the way humans interact with their telephones and the World Wide Web is nothing but a voice-based application development language designed primarily for phone transactions.

Everyone knows that XML (eXtensible Markup Language) is a standard of the World Wide Web Consortium (W3C) and is the most widely accepted, platform-independent standard for building structured documents for Web applications. Now, VXML had become a dialect of XML developed to write speech recognition scripts for voice portals.

HTML and XML applications are accessed via a graphical Web browser with display, keyboard, and a mouse while VXML applications are accessed via a voice-capable device that accepts audio and keypad input and delivers audio output through, say a telephone.

In short, VXML users interact with applications in the most natural way possible: by speaking and listening.

What can you do with VXML?

You can voice-enable a website or build next-generation interactive voice response (IVR) services and it is best suited for applications that require relatively little input from the user. Yet it can deliver highly targeted output from an HTML Web interface.

A typical application is a voice portal service, whereby callers dial a phone number to retrieve information, such as stock quotes, airline flight information, or weather from a website. Early adopters tend to use technology in this way, but VXML will gain ground for more diverse applications, such as voice-enabled intranets and contact centres, notification services, and other innovative telephony services.

Some typical voice portal applications include information retrieval, electronic commerce, telephony services, directory assistance, internal processes (security features that apply to the Web, such as firewalls and encryption, can be applied to voice applications and VXML can be used to create secure intranet applications that voice-enable internal processes, such as supply ordering, HR self-service, and corporate news), and unified messaging (for mobile employees VXML can unify voice and electronic channels, for example, by reading and recording e-mail over the phone, and originating and terminating pager messages on the phone).

What VXML does for you

Through VXML everyone can access the Web. Although the PC has been heralded as the multimedia communication portal of the future, the phone is and will continue to be important. And, with mobile phones loudly beeping their way into the communication space, VXML applications are set to proliferate.

With VXML, any telephone, even the most primitive old-style phones, can become a voice portal into the Web. A voice browser running on a telephony server interprets the input (speech or dial-pad tones) and passes it to the application logic running on the Web server. There’s no need for a cumbersome PC with Web browser and a special Internet connection. When voice-activated ‘universal remotes’ take hold, they will pass VXML content from all devices in the vicinity.

Besides, self-service applications are much more sophisticated now. For example, a VXML script could recognise part numbers, manufacturing stages, plant locations, trouble tickets, and warranty claims information—and reconcile this information to deliver updated information to support rapid marketing, engineering, or recall decisions.

Or, a VXML script could recognise street number, street name, and city for starting and ending destinations, and deliver driving directions between the two—an application that requires a very large vocabulary and interpretive processing capability.

Call treatments too can be customised on the fly. For example, if a bank customer only has a checking account, then the options for savings account and money market transactions don’t need to be offered. Or if the product manager selects the Mumbai plant, the system won’t offer options that relate only to products manufactured in Kolkata.

Moreover, applications are easy to deploy, the code can be reused among applications and platforms and because VXML is an industry specification under governance of the W3C, applications that work on one standards-compliant VXML platform will work on others as well.

With sophisticated speech recognition algorithms and large vocabularies the applications for VXML are limited only by imagination, opportunity, and market demand.

The author is national sales manager with Nortel Networks India and can be contacted at rajatt@nortelnetworks.com