Ditech Networks, whose core business involves noise suppression or elimination to support speech recognition, took the covers off of a new service called toktok™ at eComm2009 (March 3-5). The new service, which is in “invitation only” beta at this point, is a speech front end to Web services, including directories, calendars and social networks. The service toggles ‘on’ when a subscriber says the word “toktok”, even in the middle of a phone conversation. At that point a prompt solicits commands such as “make an appointment for…” or “conference in…” from the user without dropping the existing call.

Toktok elicits memories of Wildfire, Webley, HeyAnita or other early voice concierge services. Yet toktok holds the promise to deliver much more than its proprietary predecessors or even the present day unified communications and collaboration offerings from Microsoft, IBM, Cisco or others, whose solutions are only “open” to the degree that you need to employ their brand of servers for presence or calendaring.

Toktok is a packaging of services offered through mStage, which is an application platform that Ditech Networks launched in early February. Ditech will act as a services host and encourage mobile Web application developers to integrate voice into their services by employing its “open API”. The specific targets for mStage are mobile carrier. Ditech’s executives see deployment of mStage as a way for carriers to drive additional services to retain subscribers and revenue opportunities for mStage-enabled networks. Independent application and content providers could benefit as well.

The use case that captures both the spirit and substance of the service is the “voice poke”. Subscribers can use toktok to display the members of their Facebook community who are “available” and then send them a “whisper message” which they can hear during the course of a phone conversation. While I was not privy to a live demonstration, I can see great value in using this sort of utility to consult with friends about a considered purchase (furniture or some such) or for feedback on a restaurant, movie or entertainment venue.

If all of this seems far-fetched, I don’t think people could picture member of the U.S. Congress twittering away during a presidential speech.

Ditech Network has long focused on the user experience and its investment in noise suppression will play an important role in raising the probability that utterances will be understood, even in noisy environments. That will be the true litmus test, and will go a long way toward fostering user trust of the sorts of hands-free, phone-based services offered by Nuance, Microsoft/Tellme, Vlingo, Google, V-Enable and others.

Categories: Articles