Surfing the Web with Voice

An Indian startup named TringMe announced a new platform this week that it says will allow developers to easily write voice-enabled apps and widgets for the Web. But to date, voice-over-IP is a technological underachiever. Could …

An Indian startup named TringMe announced a new platform this week that it says will allow developers to easily write voice-enabled apps and widgets for the Web. But to date, voice-over-IP is a technological underachiever. Could easier development change all that?


Computer-centric VOIP should have rightfully made telephone communication obsolete by now. But it’s never really caught on in mainstream America; though Skype is popular overseas, it is less so in the US. Maybe it’s because the antics of big players like Vonage have soured consumers to the idea of IP calling. Or perhaps it’s because people just don’t realize talking through their PC is an option. Then again, perhaps we are simply uncomfortable talking at our monitors.

Making VOIP more obscure is the fact that most of us — myself included — can’t even imagine the possibilities behind it; how many of us knew the power of the smartphone before the iPhone’s app store came to show us? Easy telephony apps and widgets could mean voice-powered browsing, easier handicapped accessibility, or clicking “call” instead of “reply” on your webmail account. It could mean entering a tech support site and leaving a voicemail there instead of entering a written ticket.

And maybe it’s about time. Away from our desks, voice interaction exists all around us. Most mobile phones can do voice activated dialing these days, and RIM’s [RIMM] BlackBerrys can even do dictated emailing. RCA [VOXX] makes a voice-powered universal remote for your entertainment system. Car-maker Ford’s [F] in-dash Sync system can both play songs, and place Bluetooth calls via voice. Do we simply need a simpler platform to make the Web voice friendly, as TringMe would have us believe? Do we even want to navigate and explore the Web by speaking to our computers?

TringMe’s platform is called VoicePHP. Regular ol’ PHP is an existing Web language used to write the front end of dynamic application-rich websites like Facebook and Digg. TringMe says its VoicePHP isn’t an extension or add-on that requires any new knowledge, as would learning the APIs for other telephony Web apps like Gizmo5. PHP is the same ubiquitous language developers use now, except the inputs and outputs simply use speech instead of text on a screen. To see how it works, click here.

Since so many developers already write in PHP, it means they can easily create voice applications for their websites that can both take speech commands and talk back to users. It should make Internet calling to phones easier, too, democratizing some of the functions that took sites like GrandCentral a long time to pioneer.

What’s been the hold up? In the past, developers who wanted to make Web voice apps needed to use something called VoiceXML, a voice-tailored version of XML, or “extensible markup language.” But XML was meant to store and present data, not create applications, so it requires lots of new learning, workarounds, and hacks. PHP, by contrast, is a real programming language with real app-building power.


TringMe’s APIs will allow developers to build voice apps with front ends in Adobe [ADBE] Air, Flash, AJAX, HTML and other popular languages, meaning they can be used on smartphone platforms like Windows Mobile and the iPhone OS. (The most incredible of these technologies, Adobe Air, is the force behind powerful Web apps like eBay Desktop.) TringMe claims that its short learning curve means less time experimenting, lower cost, and less time to market — with much more breadth of use. Think of what the easy-to-use PayPal platform did for small retailers accepting online payments. Could TringMe be a stepping stone towards the proliferation of Web telephony?

TringMe’s founder, Yusuf Motiwala, seems to think so, and he knows the industry; he did stints at Lucent and Texas Instruments before starting up TringMe in Bangalore. His blog can be found here; on the right side of the page, there’s a TringMe button that lets you leave him a voicemail.

But while cars, phones, and now Web languages are getting more voice-oriented, users seem to be getting more old-fashioned about inputting data. After all — if speaking was really preferable to typing, then why are text messaging rates going through the roof, despite carriers’ price increases? Why hasn’t speech-to-text word processing software like Dragon’s excellent NaturallySpeaking overtaken typing, a task no one seems to particularly love? And why do emails seem to more often replace phone calls than snail mail?

The brains behind TringMe aren’t the only ones bullish on Web via voice though. In February, Bill Gates told an audience at Carnegie Mellon that spoken natural language search is “one of the big bets we’re making” at Microsoft [MSFT]. ChaCha and Google allow you to do voice searches by phone, but haven’t made any moves towards PC implementation.

Maybe Gates and Motiwala are right, and voice-powered Web is one of those futuristic technologies we never knew we needed. Let’s hope it’s more like wireless electricity and less like like the Segway.



About the author

I've written about innovation, design, and technology for Fast Company since 2007. I was the co-founding editor of FastCoLabs