Modifying voice response applications
Voice response application developers may write problematic applications that inherently use system resources inefficiently or are extremely large and complex. Designing and developing voice response applications requires skill and training, and it is not the purpose of this topic to address these issues.
However, here are some key guidelines for making applications efficient:
- Since applications are interpreted, using the code for anything other than basic call flow control may result in unacceptable inefficiencies. Code segments performing complex lexical or arithmetic calculations should be considered as candidates for DIPs.
- Using sub-applications results in modular programs that are more efficient. For example, a main application allows a caller to select a language (that is, a version of an application in a particular language). The caller input would then cause the main application to execute the language sub-application.
- Application scripts should be shared across channels whenever possible, and redundant code and data should be eliminated. The size of voice response applications, both code and data, affects memory usage.
- Interactions with hosts and databases should be handled efficiently, rather than being executed in a redundant manner. See Managing host interactions and Managing database transactions for details.
Besides these general guidelines, there are techniques to reduce the load for voice play. These are explained in the Avaya IVR Designer Help.
