The 2007 Speech Luminaries

Article Featured Image

Selecting this year’s winners for the Speech Luminaries awards was no easy task. We received nominations for candidates in the areas of research and development, speech science, systems engineering, system implementation, and corporate and organizational leadership. All of these individuals have had a hand in advancing the causes of their respective companies and the industry as a whole, and worked tirelessly to increase awareness and acceptance of speech technologies during the past 12 months.

We spent countless hours pouring over the nominations, making follow-up phone calls, and sending email requests for additional information. In the end, we selected a foursome that covers a cross-section of the industry as a whole, representing a leading industry association, the vendor community, and the end-user community. They have all been pioneers in the industry—in one way or another—and have spent their professional lives to this point developing and deploying solutions that have shaped how we live and communicate today.

Mark Randolph Advances Industry Through VoiceXML Forum

Mark Randolph, Ph.D., took over as chairman of the VoiceXML Forum in January, and since then the organization has expanded its technical, educational, and marketing efforts beyond VoiceXML into adjacent technologies, including speaker verification, call control and routing, Media Resource Control Protocol (MRCP), Voice over Internet Protocol (VoIP), tools, usability, and accessibility. Recently, the organization also published its first VoiceXML Forum Solutions Directory, a listing of more than 160 products and services vendors related to VoiceXML.

In addition to his work promoting and accelerating adoption of VoiceXML-based applications worldwide with the VoiceXML Forum, Randolph is also a director of the Applied Voice Input/Output Society (AVIOS) and is the director of engineering and technology at Motorola. At Motorola, where he has worked since 1995, he has toiled to bridge the gap between the spoken voice and the technological world, forming a group to develop large vocabulary speech recognition technology and natural language dialogue systems. One of the outgrowths of that effort was a prototype system for an airline travel planning service using a spoken language interface. Additionally, his team developed the first implementation of distributed speech recognition, which is now part of the 3GPP standard for voice-enabled services.

Randolph also directed the creation of the Mya voice portal. Currently he is responsible for strategy, planning, and the commercialization of technologies in the area of seamless mobility applications, content, and services.

Prior to joining Motorola, he was a member of the technical staff at AT&T’s Bell Labs. He holds a bachelor’s degree from Georgia Tech, master’s and doctoral degrees in electrical engineering from the Massachusetts Institute of Technology, and a master’s in business administration from the University of Chicago.

Michael Codini Takes Speech Beyond Voice Recognition

Michael Codini is the chief technical officer at San Mateo, Calif.-based VoiceObjects, a company he cofounded more than six years ago, and has been instrumental in leading the company’s product strategy, development, and delivery. He is the man behind the April introduction of VoiceObjects 7, a family of speech products that is being touted as the first real-time analytics solution with out-of-the-box support for common business intelligence tools and an industry-first phone application server that supplements traditional voice recognition interfaces with text, video, and Web capabilities. As a sign that the phone application server may be the next big trend in enterprise build-outs, SAP is integrating VoiceObjects phone server technology into its NetWeaver platform.

Codini’s more than 14-year background in speech and information technologies spans the areas of highly scalable/available software platforms, database processing, business intelligence, interactive voice response (IVR), voice and video portals, automatic speech recognition and speech synthesis, and the integration of all these technologies via open standards like VoiceXML.

Prior to starting VoiceObjects, Codini spent six years with business intelligence software maker MicroStrategy, four of them as head of its professional services unit in Germany and two in the company’s McLean, Va., headquarters as senior program manager for its Intelligence Server, the flagship product in its business analytics line.

He joined MicroStrategy from the Global Services group at IBM Germany, where he was in charge of data warehousing and business intelligence projects in the retail industry solutions unit. He holds a master’s degree in computer science from the Technical University of Darmstadt.

Elisa Elderbaum's Work Takes Call Center Technologies National

When Empire Medicare Services, a Medicare claims processor for healthcare facilities and medical practitioners in several northeastern states, joined four other Medicare claims operations to form National Government Services (NGS) in January, this much larger organization also gained the services of Elisa Elderbaum, Empire Medicare’s IT project manager. Since then, she has worked hard to create a paperless call center environment across all NGS’s operations and worked with the Centers for Medicare and Medicaid Services, an agency of the U.S. Department of Health and Human Services, to develop a national customer satisfaction survey standard and to create the Next Generation Desktop, which standardizes all desktop computers in Medicare-related call centers.

Elderbaum is a 20-year Medicare services veteran and has spent the last 10 years in contact center redesign. Her most recent work focused on the integration and standardization of all NGS contact center interactive voice response (IVR) systems. Those systems are being modeled after a successful call center solution that she put in place at Empire Medical Services in 2003, replacing a legacy system with an IVR that uses services and technologies from MicroAutomation, Nuance Communications, and Genesys Telecommunications Laboratories. The application improved self-service rates to nearly 80 percent (up from less than 20 percent prior to the installation), improved call processing by more than 30 percent, reduced operating costs, improved customer service, and reduced waiting times. Empire’s three offices and claims centers handled about 17,000 calls a day with just 192 ports.

Elderbaum also acts as a liaison between the technical staff at NGS—which includes the operations not only of Empire Medicare but also of AdminaStar Federal, Anthem Health Plans of New Hampshire, Associated Hospital Services, and United Government Services—and the end-user community.

Unsung Hero: CallMiner's Michael Dwyer

As vice president of research and development at Fort Myers, Fla.-based CallMiner, Michael Dwyer has served as the company’s chief architect for software application development since August 2003 and is responsible for developing and integrating state-of-the-art advances in speech recognition and analytics technologies into CallMiner’s suite of speech analytics products. He was just the fourth person to join the company, which was founded in 2002.

As head of R&D, Dwyer developed the server and database framework that underpins CallMiner Eureka!, an analytics product that the company released in March with a great deal of fanfare. Dwyer has also delivered a number of first- and second-generation products, including database back-end mining and distribution services and front-end user Web and Windows applications, from concept to customer installation. He also architected several audio search solutions that allow word and phrase searches to be conducted across more than 100,000 hours of audio in less than 30 seconds, and created an enterprise-scale audio processing and distribution system to mine thousands of hours of calls a day in parallel.

His current efforts with CallMiner include the development of next-generation speech analytics solutions that will incorporate elements of natural language processing, acoustic quality, taxonomies, and statistical matching to produce a wider range of topic identification solutions.

Prior to joining CallMiner, Dwyer developed a number of high-tech pattern recognition products—including ones for fingerprint and facial matching—for ID Software, and successfully completed the Information Management Leadership Program for General Electric. He holds a master’s degree in computer science from Case Western Reserve University in Cleveland.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues