Localization of voice prompts for the voicemail and auto-attendant system consists of two steps: a) Translation and recording of the voice prompts, and b) Creating the VXML voice scripts that assemble the sound bites recorded into meaningful sentences.

There are 487 different sound bites pre-recorded in the system in the form of .wav files, where most of them are single words.  Some information on how to localize prompts is on the Wiki:

• How to localize voice prompts
• How to record voice prompts

In order to localize the voicemail and auto-attendant systems, recording using a recording capability or studio is recommended. We expect localizations to come back to the project so that they can be maintained going forward.