VoiceXML Overview
Since its introduction, VoiceXML has become the standard technology for deploying automated phone systems. The overview of traditional technologies used to develop interactive voice response systems is given to understand the acceptance of VoiceXML by enterprises, carriers, and technology vendors.
VRU Technologies Limitations
Conventional VRU solutions are not speech enabled and upgrading to speech recognition on a traditional VRU platform is difficult. These VRU solutions do not allow the choice and flexibility necessary to meet the increasing demands of fast, quality service and a consistent experience across phone and web contact channels. These limitations of conventional VRU solutions are overcome by implementing VoiceXML as one-size-fits-all VRU solution, because VoiceXML is flexible and powerful.
VRU Development Simplification with VoiceXML
VoiceXML is a programming language that was created to simplify the development of VRU systems and other voice applications. Based on the Worldwide Web Consortium’s (W3C’s) Extensible Markup Language (XML), VoiceXML was established as a standard in 1999 by the VoiceXML Forum, an industry organization founded by AT&T, IBM, Lucent and Motorola. Today, many hundreds of companies support VoiceXML and use it to develop applications.
VoiceXML utilizes the same networking infrastructure, HTTP communications, and markup language programming model. VoiceXML has features to control audio output, audio input, presentation logic, call flow, telephony connections, and event handling for errors. It serves as a standard for the development of powerful speech-driven interactive applications accessible from any phone.
Key Business Benefits of VoiceXML
A VoiceXML-based VRU provides options when creating, deploying, and maintaining automated customer service applications. By providing the standards-based feature of VoiceXML, organizations are obtaining a number of benefits including:
-
Portability—VoiceXML eliminates the need to purchase a proprietary, special purpose platform to provide automated customer service. The standards-based feature of VoiceXML allows VRU applications to run on any VoiceXML platform, eliminating customer bound. A VoiceXML-based VRU offers businesses a choice in application providers and allows applications to move between platforms with minimal effort.
-
Flexible application development and deployment—VoiceXML enables freedom of choice in VRU application creation and modification. Because it is similar to HTML, development of VRU applications with VoiceXML is simple and does not require specialized knowledge of proprietary telephony systems. VoiceXML also is widely available to the development community so enterprises can choose between many competing vendors to find an application that meets their business needs. Increased application choice also means that businesses are not tied to the timeframe of a single application provider and can modify their VRU based on their own organizational priorities.
-
Extensive integration capability—VRU applications written in VoiceXML can integrate with and use existing business applications and data, which extends the capabilities of core business systems already in use. A VoiceXML-based VRU can integrate with any enterprise application that supports standard communication and data access protocols. By leveraging the capabilities of existing legacy and web systems to deliver better voice services, organizations can consider their VRU like their enterprise applications and fulfill business demands with an integrated customer-facing solution.
The wide array of options available allows businesses to maximize existing resources to deliver better service at lower cost.
-
Reduced total cost of ownership—The options offered by a VoiceXML-based VRU reduces the total cost of ownership in several key areas:
-
Speech capability is standard—The VoiceXML architecture directly supports integration with speech recognition, which makes a VoiceXML-based VRU a cost- effective alternative to retrofitting a traditional VRU for speech. Incorporating speech into an VRU solution increases call completion, lowering the average cost per call.
-
Lower hardware and maintenance costs—VoiceXML applications run on commonly available hardware and software, enabling businesses to save money by using equipment that they already own instead of purchasing special purpose hardware. Additionally, businesses can use the same team that handles existing enterprise maintenance to maintain VRU applications written in VoiceXML.
-
Affordable scaling— In a VoiceXML-based VRU model, application logic resides on a web/application server and is separate from telephony equipment. Businesses can save money by purchasing capacity for regular day-to-day needs and outsourcing seasonal demand to a network provider.
-
Applications for every budget—Competition between VoiceXML application developers provides a variety of VRU solutions for budgets of all sizes. Businesses pay only for necessary application features.
-
VoiceXML Use
VoiceXML is designed to leverage web infrastructure. VoiceXML is analogous to HTML, which is a standard for creating web sites. The development of voice applications using VoiceXML is simple and straightforward. Because the complexities of voice applications development are hidden from developers, they can focus on business logic and call flow design rather than complex platform and infrastructure details.
With VoiceXML, callers interact with the voice application over the phone using a voice browser. The voice browser is analogous to a graphical web browser, such as Microsoft’s Internet Explorer. Instead of interpreting HTML as a web browser does, the voice browser interprets VoiceXML and allows callers to access information and services using their voice and a telephone.
The primary components of the VoiceXML platform architecture are the telephone, voice browser, and application server. The voice browser, a platform that interprets VoiceXML, manages the dialog between the application and the caller by sending requests to the application server. Based on data, content, and business logic, the application server creates a VoiceXML document dynamically or uses a static VoiceXML document that it sends back to the voice browser as a response.
VoiceXML Development Challenges
Despite the robustness and broad acceptance of VoiceXML as the standard for voice applications, there are a number of challenges that developers face when deploying complicated systems, including:
-
Requirement for dynamic VoiceXML— Many applications need to dynamically insert content or to base business logic on data available only at runtime. In these cases, the VoiceXML must be dynamically generated. For example, an application that plays a Good Morning / Afternoon / Evening prompt, depending on the time of day requires VoiceXML to be dynamically generated.
-
Voice paradigm versus web paradigm—There are many systems designed to manage dynamic web content or to automatically convert web content to other formats (such as for wireless phones). These systems, however, are not adequate for voice applications due to the fundamental difference between a voice application and a web application. A web page is a two-dimensional, visual interface while a phone call is a one-dimensional, linear process. Converting web content to voice content often produces voice applications with flat user interfaces.
-
Browser compatibility—Due to ambiguities and constant improvements in the VoiceXML specification, no two commercially available browsers accomplish various functions in exactly the same way. Developers must understand the variations between browsers when coding VoiceXML to ensure compatibility.
-
Stateless nature of VoiceXML—Like HTML, VoiceXML is a stateless mark-up language. For applications that require the maintenance of data across a session, for example, account or transactional information, or phone call, VoiceXML does not suffice.
-
Complicated coding—Despite VoiceXML’s function to simplify voice application development, the process of coding an application with dozens or hundreds of possible interactions with a caller can become quite complex.
-
Limited back-end integration—Enterprise applications rarely operate without any connection. VoiceXML does not natively support robust data access and external system integration.
-
OAM and P requirements—Operators of large-scale voice applications have significant requirements for administration, management, logging and (sometimes) provisioning. VoiceXML does not natively support most of these functions.
-
Reusability—The larger a web or voice application becomes, the more critical reusability becomes. This is even more definite in dynamic applications. VoiceXML only provides the interface for a voice application. VoiceXML does not provide common application functionality into configurable and reusable building blocks.