Dave Nuckols, Director of Microsoft Strategic Alliance, ScanSoft

NewsBlast: The official launch of the Microsoft Speech Server will undoubtedly generate greater awareness of the speech industry by enterprise customers large and small. As a veteran of the speech industry, what advice does ScanSoft offer to organizations that are evaluating speech for the first time?
Dave Nuckols: First of all, just go for it! ScanSoft has hundreds of customers around the world and has delivered incredible business value to an array of organizations. Aetna, Bank of America, Boeing, United Airlines and Verizon have collectively generated hundreds of millions of dollars in savings and increased caller satisfaction dramatically. Speech recognition is clearly mainstream in the Global 1000 market. The ScanSoft - Microsoft partnership is designed to further accelerate adoption.

The single most important piece of advice we give our customers is to focus on the speech application design and development process. The only thing the caller hears is the performance of the application. The performance of the application is what unleashes the business value. So, be sure to partner with a company that has significant experience developing speech applications in your specific industry. Ensure your partner has relevant proven user interface design skills. Be sure they know how to apply core technology for optimal application performance. In short, make sure they can deliver powerful speech applications.

The second piece of advice we give customers is to develop a speech strategy. ScanSoft regularly engages C-Level executives who want to truly transform the call center experience. A speech strategy includes a multi-business unit application roadmap and a coordinated telephony brand or voice identity. Did you know one of our large banking customers takes over half a billion phone calls per year? That means they make an impression on their customers 500 million times via the telephone. That's incredible. Savvy executives recognize the strategic importance and are investing in speech recognition for the next generation call center. Interestingly, the next generation call center is tightly integrated to the very same Web infrastructure and utilizes telephony, multimodal and Web interfaces.

NB: What benefits does ScanSoft offer to enterprises deploying speech services on Microsoft Speech Server 2004? After all, Microsoft has a lot of partners.
DN: ScanSoft has the broadest speech technology portfolio in the world. We have built more speech applications in more industries in more languages than any company in the world. We have a proven track record of customer success stories and benefit from a wealth of industry specific best practices. That expertise is designed into our products, applications, and solutions.

Microsoft selected ScanSoft as a strategic partner in part for these reasons. The Microsoft Speech Server supports our Speechify 3.0 TTS bundled with the platform. Open Speech Recognizer 2.0 speech recognition is available on Microsoft Speech Server. Our core speech technologies are optimized for superior application performance and offer scalability, accuracy and support for natural language dialogs that result in a more natural, and conversational dialog with callers.

Interestingly, our professional services team already built one of the largest and most complex applications yet on the Microsoft Speech Server. So no one is better positioned to deliver outstanding applications and solutions on Microsoft Speech Server than ScanSoft.

NB: Speech applications and particularly packaged applications are hot topics right now. Can you describe your application strategy relative to Microsoft Speech Server?
DN: ScanSoft maintains a proprietary knowledge database of more than 300 industry-specific applications honed from our consultative STEP program and Professional Services experience. We utilize application design best practices on all deployments and plan to do so on the Microsoft Speech Server. We launched our SpeechPAK strategy in 2003 including SpeechPAK for Healthcare, SpeechPAK for Directory Assistance, and OpenSpeech Attendant. Needless to say, Microsoft and ScanSoft plan to collaborate on packaged applications.

NB: How can other Microsoft partners begin working with ScanSoft?
DN: Great question. ScanSoft partners with others in the Microsoft ecosystem. If you are a "Microsoft partner" then you are already a "ScanSoft partner" too. ScanSoft's portfolio of complementary technologies and applications are available on the Microsoft Speech Server and can be resold by qualified ScanSoft partners. Nearly all of our business goes through channel partners even when our sales team is involved in creating new sales opportunities, and it will be the same within the Microsoft ecosystem.

Partners may come to us directly for software and sub-contracted professional services support or they may obtain our software indirectly through authorized distributors such as those selling Intel Dialogic Boards and the Intel or Intervoice Telephony Interface Manager (TIM) - all components which come together to complete a Microsoft Speech Server implementation.

NB: Microsoft does offer its own ASR, so it's not just an IVR platform. So which is it for you two, competitors or partners?
DN: Microsoft is a valued strategic partner and we expect great things from this relationship. We work closely in all realms from technology to joint sales and enjoy a preferred status for application and user interface design services.

Microsoft is also a large end-user customer of our services across its own automation projects. We were delighted that they chose us to design and develop their first major customer-facing deployment of Microsoft Speech Server within their own call centers.

You are right to say "Microsoft isn't just an IVR platform." They do offer a speech recognition engine, but there is additional value in Microsoft's development environment. The Speech Applications SDK turns Visual Studio.NET into a very robust environment for soup-to-nuts IVR/ASR development, and that's true regardless whose ASR you use. Implementation - both in human factors and our distinctive speech project methodology - is still the key to a successful project. And the strength of their development environment allows us leverage specialized know-how with less wasted effort.

And as you pointed out at the beginning, Microsoft's entry into this market will shake things up and bring many new enterprises into the market for the first time. Not just because of the IVR and Tools but because they have a full software stack that goes deep into the data center with web, application, messaging, and database infrastructure elements just to name a few. So bringing this all together will drive down the cost for deploying speech. So, for us, Microsoft is not just a partner but rather a strategic partner with whom we can reach broad new markets for our applications, technologies, and services.

NB: Okay, then, so tell me: Why would a Microsoft partner or end-customer insist on ScanSoft's Open Speech Recognizer (OSR) for the Microsoft Speech Server 2004?
DN: In addition to Microsoft's ASR engine, our OpenSpeech Recognizer is available on the Microsoft Speech Server 2004. OpenSpeech Recognizer is a field proven speech recognition engine, with demonstrated world-class accuracy and scalability on the largest deployments in the world. One of our large telecommunications customers takes more than 1.5 million calls per day. Furthermore, OSR offers advanced capabilities to improve the caller experience and boost return-on-investment. For example, the unique SpeakFreely capabilities allow speech applications to support natural language responses to open-ended questions, streamlining calls while expanding the range of tasks that can be automated successfully. In conjunction with OpenSpeech Recognizer we also offer our SpeechSecure voice authentication engine for remote biometric verification by voice print.

Enterprise and carrier customers around the world are already taking advantage of OpenSpeech Recognizer on many leading IVR platforms and use our complementary tools such as our OpenSpeech Insight (OSI) reporting tool to further analyze and refine their speech applications. Enterprises who rely on the expertise of ScanSoft professional services for speech application design will find OpenSpeech Recognizer a compelling choice for deploying speech services on the Microsoft Speech Server.

NB: Any final thoughts?
DN: We are excited about the prospects for speech and the growing adoption for our technologies, applications and services from enterprise and carrier customers around the world. The entire SpeechTEK team has been a huge service to this industry and we're confident that both the SpeechTEK publication and conferences will continue to aid in educating enterprise and carrier organizations on the benefits of speech for their business.

Note: ScanSoft is a Gold Sponsor of AVIOS~SpeechTEK 2004 and a Premier Sponsor of Microsoft Speech Server launch. Visit ScanSoft in Booth #400 and in Microsoft's booth #928.

Partners and customers interested in learning more about ScanSoft products and services for Microsoft Speech Server 2004 are invited to contact Dave Nuckols at 952-931-9417 or dave.nuckols@scansoft.com.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues