SPEECH TECHNOLOGY: Goes Mainstream

author-image
DQC News Bureau
New Update

Since the dawn of the information age, researchers have labored to build
applications that will enable people to interact with computers simply by
talking. But building computer systems that respond instantly and accurately to
the spoken word has proven to be much more difficult than most people
anticipated.

Advertisment

Because of the sheer complexity and variety of human speech, only the largest
organizations have been able to afford the time and expense required to build
sophisticated speech-enabled applications. As a result, speech recognition
software, that can offer much more than "Press one or say one"
functionality, has been something of a rarity.

With the launch of Microsoft Speech Server 2004, the world of speech
recognition software is poised to undergo an important change. Introduced by
Microsoft Chairman and Chief Software Architect Bill Gates during his opening
keynote address at the joint VSLive! and AVIOS~SpeechTEK event in San Francisco
recently, Microsoft Speech Server is a standards-based, integrated speech
platform that will deliver the business benefits of speech-enabled solutions to
both large-scale enterprises and mid-size companies.

The release of Microsoft Speech Server marks the shift of speech recognition
from a niche technology to mainstream application that can help companies reduce
costs, raise productivity, increase customer satisfaction and take advantage of
new opportunities for generating revenue.

Advertisment

BEGINNING OF SPEECH TECHNOLOGY

First made available in July 2003 with the release of the beta 1 version,
Microsoft Speech Server is already playing a significant role in helping
independent software development companies create mainstream speech-enabled
solutions.

Currently, more than 60 companies are working with Microsoft to develop
packaged speech applications and services through the Microsoft Speech Partner
Program. Companies such as Jet Blue and Grange Insurance, as well as the New
York City Department of Education, are taking advantage of speech-enabled
solutions built on Microsoft Speech Server to streamline operations and deliver
better customer service.

HOME-GROWN
SPEECH TECHNOLOGY

Tata
Consultancy Services (TCS) is demonstrating Microsoft Speech
Server-based solutions at SpeechTEK. TCS was the first Microsoft
technology partner to implement Microsoft Speech Server into a live
customer environment.

Built
for Grange Insurance, the speech-enabled solution was created to
enable callers to retrieve billing details of a policy 24 hours a
day, seven days a week. When it was deployed, the system was built
to handle 20% of policy inquiry calls, with the expectation that it
would scale, over time, to handle 80% of those calls.
According
to TCS Development Center Manager Nitin Hanjakar, the success of the
Grange Insurance application has led TCS to explore a wide range of
potential applications based on Microsoft Speech Server. At
SpeechTEK, TCS showcases round-the-clock-billing-inquiry
application, which allows callers to get policy or mortgage billing
details over the phone. The solution not only allows customers to
access key information 24 hours a day, but it reduces customer
service representative workload significantly, enabling them to
focus on cross-selling and up-selling opportunities.
Advertisment

Along with the Bill Gates’ speech announcing the general release of
Microsoft Speech Server, SpeechTEK Spring 2004 includes demonstrations of speech
technologies and solutions built by companies that have worked with Microsoft
during the development of Microsoft Speech Server and as part of the Microsoft
Speech Partner Program.

Among the companies at SpeechTEK Spring 2004 are such leading communication
technology companies as Tata Consultancy Services, Intel, Intervoice, ScanSoft,
Envox Worldwide, Voice Automation, Gold Systems, Redmond Software and CYGCOM.

WORLDWIDE DEVELOPMENT

Intel Communications Group has worked closely with Microsoft to develop a
reference architecture that enables Microsoft Speech Server to interoperate with
existing telephone networks. It is providing two key pieces of that reference
architecture: the telephony board, which is a piece of hardware built by Intel
that allows voice-enabled solutions based on Microsoft Speech Server to connect
directly with existing telephone systems, and the Telephony Interface Manager
(TIM) which is a layer of software that simplifies integration of the telephony
board with Speech Server.

Advertisment

Intervoice, a Dallas-based company that has been developing voice solutions
since 1983, has also worked with Microsoft to develop a TIM. At SpeechTEK, the
company presented how the Intervoice TIM can help turn Microsoft Speech Server
into a robust, feature-rich voice automation platform that integrates into
virtually any call center or telephony environment.

ScanSoft, a global leader in speech technologies, applications and services,
is another Microsoft partner that is excited about the potential for growth in
the market. At SpeechTEK, ScanSoft demonstrates its Change of Address packaged
application that automates the way call centers collect address information.
Change of Address can be used to start or stop service for a utility or to
forward a magazine subscription to a new address.

SPEECH-ENABLED SOLUTIONS

Microsoft and Voice Automation announced a strategic relationship aimed at
voice-enabling Microsoft Business Solutions CRM last October. Voice Automation
for Microsoft CRM allows remote workers to access and update enterprise sales
data without a keyboard.

Advertisment

With the Voice Automation solution, employees who are on the road will be
able to use a phone to provide spoken commands that allow them to navigate
through Microsoft CRM.

Voice Automation for Microsoft CRM offers simplified, secure login from
either a land or mobile phone. At SpeechTEK, Voice Automation showcases key
features of the solution including the ability to retrieve and respond to
e-mail; access contact information details such as telephone numbers and driving
directions; and update lead status.

Colorado-based Gold Systems demonstrated two solutions based on Microsoft
Speech Server. The first is aimed at the insurance market, and includes a set of
pre-packaged speech modules that provide self-service access for policyholders.
The reusable modules, which can be easily customized to meet the specific needs
of individual companies, can significantly reduce call center costs and increase
customer satisfaction.

Advertisment

Another key participant in the Microsoft Speech Server Program that
demonstrated a new product at SpeechTEK is CYGCOM. Based in Canada, CYGCOM will
showcase its SpeechWatch Server, which monitors critical elements of system
performance. Built for enterprise deployment, SpeechWatch Server provides IT
professionals with reliability in a fully supported platform for distributed
speech applications.

Redmond Software was also at SpeechTEK with its speech-driven auto attendant.
The system, which can accurately and efficiently transfer inbound calls to the
appropriate person or department within a company without operator intervention,
includes tools to get names automatically from Microsoft Active Directory,
Microsoft SQL Server or from text files.

Other Microsoft Speech Partner Program members slated to unveil
speech-enabled solutions built on Microsoft Speech Server include Edify Corp,
Envox Worldwide and GM Voices. Partners that are providing system integration
services and customer application development services include Enterprise
Integration Group, HeyAnita, HTI Voice Solutions, InfoSys and Wipro
Technologies.

Advertisment

Microsoft Press Pass