IVR/Voice Portals
Datapoint provides open standards based telephony self service platforms (voice portals) that can support a variety of technologies which include DTMF recognition (touch tone), speech recognition, voice verification (voice biometrics) and text-to-speech.
Voice portals embody open standards like VoiceXML and CCXML to provide open and extensible applications that can be easily deployed and maintained utilising existing or extended web systems infrastructure.
Range Of Applications
The voice portals support a wide range of applications including:
- Intelligent call steering
- Customer identification and verification (ID&V)
- Self service information
- Bill payment
- Transactional applications i.e. booking a flight or transferring funds
- Proactive contact and event triggered communications
They can support traditional telephony as well as Voice over Internet Protocol (VoIP) through either H.323 or Session Initiation Protocol (SIP). They also support hybrid installations to fit with migration strategies.
Anatomy Of A Voice Portal Solution
A typical Voice Portal platform comprises of four main components:
The Telephony Gateway
This provides the telephony interface which provides IVR functionality such as creating and terminating telephony sessions as well as interfacing to speech recognition services. The Telephony Gateway hosts the VoiceXML browser which interacts with web server infrastructure for business logic and data access. This provides callers a call control and voice dialog session under control of ccXML and VoiceXML 2.1 markup languages.
The Operations, Administration & Maintenance (OA&M) Component
This is used to administer licensing and registration of application resources and services. The OA&M platform is also responsible for actively monitoring the Telephony Gateway(s) collecting errors, alarms, logs and session details. Typically the OA&M platform also generates reports based on collected Call Detail Records (CDR), Session Detail Records (SDR), and Application Detail Records (ADR) as well as supporting access to collected data via schema and database access methods. System reporting can extend from simple call volumes/port utilisation to complete tracking and audit facilities of the end to end customer experience.
The Application Server
The Application Server that the application runtime is hosted on are typically industry solutions i.e. IBM WebSphere, BEA Weblogic, and Apache Tomcat on various operating systems. Applications can be written in any number of ways based on the standard VoiceXML 2.1 and CCXML 1.0 specifications. Developers can write applications by hand-coding VoiceXML and ECMA script, by using a VoiceXML certified environment to write static VoiceXML, or they can generate dynamic VoiceXML to drive the platform.
Most Voice Portal platforms provide an application development environment. Since the applications are standards-based, there are also several third party development environments that can support the full lifecycle of designing a speech self service application.
The Speech Server
This is a generic term used to describe the server that hosts the speech recognition, verification or text-to-speech resources. Typically these resources are engaged using the Media Resource Control Protocol (MRCP). The Speech Server typically supports speech technologies ranging from standard directed dialog speech recognition to “open speech” or “say anything” dialogs.
In House ExpertiseDatapoint has extensive experience in the unique design challenges that effective speech solutions demand. Our methodology for prioritising applications that yield a strong ROI and our ongoing iterative testing process ensures that the final solution meets time, budget and end user expectations.
Partners
