Todd Lebor provides a third way to play the hands-free communication trend. In the third part of this three-part series, he argues the merits of speech-recognition software and looks at the major players, SpeechWorks and Nuance. Looks like carpal tunnel syndrome may soon be a thing of the past.
 |
|
|||||||
By
Speech Recognition
In 2001: A Space Odyssey, Stanley Kubrick created an omnipresent, voice-activated mainframe computer named HAL that ran the spaceship and wanted to run much more than that. One scene even insinuated that HAL was capable of reading lips. When 2001 debuted in 1968, such technology was mere fiction. Today, it's not that far from the truth. There are a number of public and private companies developing speech-recognition technology and yes, even lip-reading technology.
Before getting started, I want to differentiate between voice recognition and text-to-voice technology. Both are frequently referred to under the umbrella of "speech-recognition technology" and they often work in conjunction. But the two technologies function differently.
Speech-recognition software translates a voice into audible keystrokes. It turns hundreds of thousands of words with hundreds of accents into ones and zeros. Once the voice is digitized, it commands a computer to do things.
Text-to-voice software is just what it sounds like: it translates the written word into speech and then provides audio responses. Even though these responses often sound like words linked together with the precision of a blind butcher, a survey by a major voice-recognition software maker revealed that 80% of callers prefer voice interaction to older systems like touch-tone systems.
Speech-recognition is often referred to as the next "killer app" -- one of those technologies that will be integrated into everyday life for everyone everywhere. Voice commands will replace keyboard strokes, opening the Internet up to the "technologically challenged" as well as some of the physically challenged, such as those who are blind.
Speech-recognition technology is already used in many industries. To the dismay of most, speech-recognition programs have replaced humans on nearly every customer-service help line. Say "yes" after the tone if you hate those robotic programs. Beep.
But things are changing. All sorts of factors are making our dreams a reality. Computer processing speeds are now fast enough to enable "conversational" interaction between people and computers. Voice networks and data networks are converging and speech-recognition software is breaking through.
For example, just the other day I called United Airlines to check on a flight's arrival status. I never spoke to a human. But, here's the kicker: I never pushed a button on my phone either. The entire transaction was verbal and remarkably conversational. At one point, the computer (using a prerecorded human voice rather than a digitally generated voice) even asked me, "Let me be sure I understood you correctly, did you say..." I was amazed at the precision and well-designed interaction of the system.
The fact of the matter is that these productivity-enhancing programs are not going away. They are here to stay and they are expanding. Processing speeds and software enhancements have reopened investors' eyes. E-commerce move over, here comes V-commerce.
Two of the industry leaders in speech-recognition technology are Boston-based SpeechWorks International (Nasdaq: SPWX) and Nuance Communications (Nasdaq: NUAN) in Menlo Park, California. Both experienced strong revenue growth (105% and 148%, respectively) over the first nine months of 2000 and there's no end in sight.
SpeechWorks provides over-the-telephone automated speech recognition (ASR) solutions. It went public in August to much fanfare. SpeechWorks sports a $2.8 billion market capitalization based on $19.7 million in revenue and a net loss of $21.3 million for the nine months ended September 30, 2000.
On October 10, SpeechWorks announced that Yahoo! (Nasdaq: YHOO) would begin using its Speechify Text-to-Speech engine to enable Yahoo! Mail users to check mail over the phone. Shortly afterward, America Online (NYSE: AOL) signed a licensing agreement to use its software for the AOL by Phone service. What's next -- a totally voice-activated portal? Yep.
One of the hottest companies raising venture capital right now is doing just that. Tellme Networks is trying to help businesses build, run, and promote voice-activated phone applications on its VoiceXML (VXML) network. Basically it is building a voice portal. In record pace, Tellme has raised $238 million in three rounds of financing in only 14 months. Not an infrastructure play, Tellme licenses its software from Nuance.
Nuance is also in the ASR business. It offers server-based software that allows phone users to interact with voice-driven telecommunications, enterprise and Internet systems. Nuance reported $34.4 million in revenue and a net loss of $17.5 million for the nine months ended September 30, 2000. As of September 30, it had $93 million in cash and short-term investments (thanks to an April IPO) and currently boasts a $3.3 billion market cap.
Nuance just signed up Ask Jeeves (Nasdaq: ASKJ) to combine the natural language, question-answering technologies of Ask Jeeves with the speech recognition technology of Nuance. Nuance products are used by major consumer-oriented companies like Charles Schwab (NYSE: SCH), Fidelity, Sprint PCS (NYSE: PCS) and General Motors' (NYSE: GM) OnStar service.
This series focused on different ways to participate in the hands-free communication trend. All three concepts may soon converge as a wireless headset employs Bluetooth to communicate with a voice-controlled handset. Want to get a headstart on the trend? We've introduced you to the major players that have to come together to make it happen.
RSS Headlines
Fool UK
