Price InterActive Selects the Intel Voice Portal Platform for Speech-Enabled ASP Infrastructure

Business Wire, Sept 27, 2000

Business Editors/High-Tech Writers

Continuous Speech Recognition Technology Enables Scalable,

High-Density Solution for PriceInteractive

PARSIPPANY, N.J.--(BUSINESS WIRE)--Sept. 27, 2000

Intel's Dialogic subsidiary today announced that they will deliver their voice portal platform with continuous speech processing (CSP) capabilities for SpeechPort, a voice-enabled Internet service from PriceInteractive, a leading speech application hosting provider.

After an intensive technology evaluation, the newly launched platform from Intel proves to have unmatched scalability and density plus interoperability with the SpeechWorks software platform.

Voice-Enabled Internet Hosting

SpeechPort service allows e-Businesses, content providers and portals to deliver speech-activated Web content through a natural language interface. By providing the hardware and management infrastructure in an ASP model, PriceInteractive enables its customers to concentrate on the value of their applications instead of on acquiring and managing infrastructure.

Choosing the Platform

Before it could implement the SpeechPort service, PriceInteractive needed to choose a speech platform. After a complete technical review of alternatives from several different suppliers, PriceInteractive chose CSP-enabled hardware from Dialogic with the SpeechWorks speech enablement platform from SpeechWorks International.

"There were three key reasons why we chose the Dialogic platform," explains Dan Price, president and CEO of PriceInteractive. "The first was scalability. We needed a system that could quickly and easily grow with our business. The Dialogic hardware gave us the flexibility to handle our rapid growth and easily manage our business.

"Density was another issue," Price adds "The Dialogic solution gave us the highest density available, which suited our needs today and gave us room to expand. High density also improves price/performance. We also liked the close partnership between Dialogic and SpeechWorks International," Price continues.

"The Dialogic CSP-enabled hardware works very well with the SpeechWorks software we chose. Also, the close partnership between Dialogic and SpeechWorks International makes us most comfortable with the overall solution."

Dialogic CSP

Dialogic CSP technology enables developers to build standards-based speech recognition applications like Internet voice portals with maximum flexibility, scalability, and density at lower cost than alternative platforms.

CSP optimizes the performance of host-based speech recognition engines by streaming preprocessed voice data between the telephony boards (analog, T-1, E-1) and the host computer's central processing unit (CPU). Because CSP is on the Dialogic boards used for front-end signal processing, the host system is better equipped for speech recognition. Dialogic has incorporated CSP onto its JCT series of voice processing boards, eliminating the need for additional, dedicated speech hardware.

"CSP technology is ideal for the kind of next-generation solution PriceInteractive was looking to build," explains John Amein, Dialogic vice president of marketing. "It is optimized to provide the scalability and flexibility these solutions demand. Also, the high density of our boards makes it possible to build a solution that uses fewer boards and less square footage -- which saves our customers money."

The Dialogic speech platform is delivered on the Intel(TM)ISP2150 high-performance 2U rack mountable chassis, enabling reliable, flexible, and high-density solutions in a small footprint. Scalable configurations can support 36, 42, 60, and 96 CSP ports, for a high-density solution that improves price/performance. The platform integrates Dialogic telephony and network interface boards, providing voice and global network interface options.

CSP consists of software and firmware features for Dialogic communications boards. It enhances existing speech technologies with board-level firmware that processes real-time voice signals that identify human speech input and present it to the host platform for speech recognition. The real-time functions include echo cancellation and voice activity detection (VAD). CSP also provides algorithms that handle voice quality and transmission speed issues. This approach offloads host platform resources for more complex speech recognition tasks like analyzing and recognizing the speech input in support of the application.

CSP supports important application-friendly features like the ability for callers to interrupt speech prompts by speaking over them -- known as barge-in functionality.

Knowledgeable callers can take control of the conversation and complete the interaction more quickly, delivering a more pleasurable experience for the caller. Barge-in also makes more efficient use of the platform for the hosting company, saves host system resources, improves system usage, and can reduce phone charges by shortening the transaction duration.

SpeechWorks Software

Enabled by Dialogic CSP, the SpeechWorks software gives PriceInteractive the advanced speech features it needs. These include speaker-independent recognition of continuous speech, a large vocabulary, real-time vocabulary generation from text, database integration, adaptive channel normalization, and support for multiple languages.