HP Calls on PipeBeach for VXML

Hewlett-Packard Wednesday said it has acquired PipeBeach, a Stockholm-based developer of VoiceXML products for an undisclosed amount.

Based on the Extensible Markup Language (XML), VXML provides a high-level programming interface to speech and telephony resources for application developers, service providers and equipment manufacturers. The World Wide Web Consortium (W3C) published the second version of its specification back in January of this year.

“Their software products and expertise were a real match to layer on top of our OpenCall media platform for more complete solution,” Ed Verney, director, interactive media platforms, for HP’s OpenCall business unit told internetnews.com.

The Palo Alto, Calif.-based computer and printer maker is working with mobile service providers to adopt its Opencall product, which has location-based services for in-call advertising and real-time pre-paid billing.

HP has partnered with PipeBeach for more than a year; Verney said the two cooperated on building a voice portal for a French telecom.

PipeBeach’s speechWeb bridges the telephone network and standard Web servers, using XML combined with speech recognition and text-to-speech applications. It lets mobile phone users interact with VoiceXML Web pages by asking for information in words rather than navigating a text menu on the mobile phone screen. Its speechWeb Portal Framework lets service providers create voice-enabled portals, so that a user can call a single telephone number and pick from a variety of information services.

The buy is expected to payoff in the next four to five years. According to SRI Consulting, spending on voice technology — including consultancy and systems integration — will jump from $20 billion in 2005 to $40 billion in 2007.

HP said it will add and rebrand PipeBeach’s technology to its own OpenCall product line.

“It differentiates us that we have the complete value chain,” Verney said. “We can take what they have and what we’ve invested in, bring it into one engineering organization and optimize it.”

Verney said the acquisition also includes a lot of industry influence. Scott McGlashan, CTO of PipeBeach, was co-chair of the W3CC working group for XML. McGlashan will become CTO of the OpenCall interactive media platforms unit. All 17 PipeBeach employees will become HP staffers based in Stockholm.

The Evolution of VXML

The inception of VoiceXML lies with an AT&T project dubbed
Phone Markup Language (PML). In 1995, AT&T created an XML-based dialog
design language intended to simplify the speech recognition application
development process within PML. With the reorganization of AT&T, teams at
AT&T, Lucent and Motorola continued working on PML-like languages. Following a W3C conference on voice browsers in 1998, AT&T, IBM , Lucent and Motorola — all of which were developing speech-based markup languages, created the VoiceXML Forum to pool their efforts and define a standard dialog design language for building conversational applications.

VoiceXML Forum released VoiceXML 1.0 to the public in 2000, and then submitted the specification to the W3C. The specification slots into the W3C’s work on the Speech Interface Framework, which would allow people to use any telephone to access appropriately designed Web-based services.

VoiceXML is similar to the recently announced Speech Application Language Tags (SALT) specification which has also been submitted to the W3C. SALT is a set of light-weight extensions to existing markup languages, particularly HTML and XHTML, that enable multimodal and telephony access to information, applications and Web services from PCs, telephones, tablet PCs and wireless personal digital assistants (PDAs). However, VoiceXML focuses on telephony application development while SALT is focused on multimodal speech application development (which, for example, would allow a PDA user to fill out a form using both voice commands and stylus, whichever is more convenient).

While the two are different, the specifications do share similar goals and
may eventually converge. In fact, SALT uses key components of the Speech
Interface Framework, including SRGS and the SSML.

Editor’s note: Internetnews.com editor Thor Olavsrud contributed to this report.