Nuance’s Ricci on speech in the enterprise

Massachusetts speech software house Nuance Communications Inc. is trying on a new identity, after having been known as ScanSoft, a name more associated with the firm’s roots in imaging, for all of its life.

After the former ScanSoft completed its acquisition of California’s Nuance, which focused on the voice-automation market, it took Nuance’s name last month, finally putting its brand emphasis on the growing speech part of its business. Although formed in 1992 as an imaging company, ScanSoft has been moving into speech technology since its 2001 acquisition of that portion of Belgian Lernout & Hauspie’s business. In the next few years it picked up speech-processing telephony and voice-control groups from Philips of the Netherlands, as well as SpeechWorks International, LocusDialog, Telelogue, Brand & Groeber Communications and Rhetorical Systems, all companies with speech expertise. In 2005, it closed acquisitions of ART Advanced Recognition Technologies and Phonetic Systems before making the Nuance buy.

Overseeing the company’s acquisition strategy and establishment as a speech technology force has been Chairman and Chief Executive Offer Paul Ricci, who joined in 2000. Ricci chatted with IDG News Service recently about the role speech technology plays in enterprise IT.

IDGNS: What are the biggest challenges that you see facing enterprises that want to work with speech technology?

Ricci: Maybe the way to answer that is to talk about what are the things motivating enterprises to work with speech. Number one is the desire to improve the productivity of their contact and call centers. … The second is to ensure quality and customer satisfaction among their customers that they’re having difficulty delivering today, in part because the ability to attract the quality of people they need to call centers is difficult. And the third is, the desire to increase the access that the customers have to the ultimate information that the company is trying to provide, whether it’s a bank account, or information about a particular service.

What enterprises need to do, to achieve those goals with speech, is no different than what they need to do in any large IT implementation. They need to approach it with clarity of goals, with a realism about the timing of the goals, with some level of prudent phasing of the implementation, with a commitment to improve the quality of the implementation over time. And I think it’s important that they find a trusted partner to help them with the implementation and the technology, because we are after all talking about a solution that affects their most important asset, which is their customers.

IDGNS: What are some of the biggest technical challenges facing your company right now? What are some of the things you’re working on?

Ricci: Probably the largest area of investment in speech today is in moving beyond the problem of word or phrase recognition to more natural language processing, that is, giving the user the ability to have a more natural dialogue with the call center. And the reason you want to do this is first of all, it increases the likelihood of a successful outcome for the call, because we don’t tend to talk in machine-like ways, and secondly because it’s a more satisfying experience for the customer.

This investment area is also playing into another interest that companies have, which is using their voice service as an extension of the branding and positioning of the company. Companies have a particular view about how they want to be perceived and we’re working with those companies to help them think about how their voice service can be consistent with that. You can imagine very different companies, staid or edgy companies, having a very different approach to that. We’re also investing similarly in speech synthesis for the same reasons, making it more expressive and more able to convey emotion in its reading. There’s an enormous appetite right now in enterprises for customized voices, that is, taking a personality or a particular style of voice and having a voice built for them, a synthesized voice, that reflects who they want to be.

IDGNS: What’s involved in implementing speech technologies around the world, for customers who need different languages?

Ricci: The core speech technology serves most languages. There are some particularities. Of course every new language has to be developed and the so-called language models have to be developed based on the sounds of that language. One real distinction that Nuance brings to this business, in part because of the L&H acquisition and in part because of the acquisition of the Philips business, is a very broad language coverage and also a fairly global perspective on the business. And that’s turned out to be important for us …when you work with a large cell-phone manufacturer like Nokia, or you work with a large automotive manufacturer, not only are they not interested exclusively in the U.S., or Germany, or English markets, they’re increasingly interested primarily in the growth markets that are elsewhere.

IDGNS: How do you view Microsoft as a competitor in speech?

Ricci: I think Nuance’s solutions have gained acceptance in large enterprise deployments, and I think Microsoft has achieved less acceptance in that area. Where I think Microsoft is most likely to be successful is in the recruitment and cultivation of their robust developer community in building applications that are linked to Web apps that those developers are doing. They’ve not been a particularly strong force in call-center automation to date, and I don’t really see them as having a great deal of traction there in the short run.

IDGNS: How about IBM?

Ricci: Of course, IBM has a long tradition of speech research, and they have some deployments, particularly through their own global services organization. But not so much outside of that.

IDGNS: What are your biggest business challenges right now?

Ricci: In the next year and next several years we’ve set our focus on a number of markets: the enterprise network market, the mobile speech market, and the health-care market. … And you might expect that puts operational challenges on a company; ability to recruit and staff in accordance with what’s necessary. I think the second area that’s a challenge is, this is a technology that needs and will benefit from continued very high levels of investment and we need to continue to make those investments. … But the single most important challenge for us at the end of the day is focusing on the success of our large deployments. Because these are essential, contact centers are critical to large companies, and we can’t afford for any of them to have a significant failure. And so an enormous amount of the company’s resources are aligned around ensuring the success of big customer deployments.