Note: You configure SAA using Web Console.
Speech Auto Attendant (SAA) is a speech-enabled application that allows users to place calls to people quickly and efficiently by speaking their names. In addition to placing calls by name, users can say a department name or telephone number. A tutorial introduces users to the system features, and voice-based help is available to answer questions.
The SAA feature uses an automated attendant to route incoming calls based on spoken commands. Typically, you say the name of the person you want to speak to and the system transfers your call to the requested party.
SAA also supports up to 15 departments. When a callers states the word "Department" (to select a department), the caller hears, "The departments are..." followed by a list of department names. Upon hearing the list, the caller states the department name and is transferred to the department. If the Barge In command has been enabled, the caller can state the department name without having to listen to the list and be transferred directly. If no department names have been configured in the Speech Auto Attendant, the system states "there are no departments".
Note: When callers say “Department”, NuPoint UM uses the Text To Speech (TTS) option to list the departments even though you may have recorded the department names. Callers can also ask for a particular department (for example, “Sales”). In this case, NuPoint confirms the department name using the recorded name, if available.
SAA can be installed as a standalone application (without any of the voice mail features included in NuPoint UM), or it can be installed along with NuPoint UM. When installed with NuPoint UM, SAA stores up to 4 telephone numbers per person and provides the user-configurable option of selecting which number will be called. When installed as a standalone application, SAA stores 1 number per person, and this number is configurable only by the domain administrator.
If you are planning to use the Speech Auto Attendant feature with the DMG integration, then you need to set the "Voice Activity Detection" option to OFF to ensure that the Speech Auto Attendant functions properly. See the NuPoint Unified Messaging Technician's Handbook for information about configuring DMGs.
There are three types of users for the Speech Auto Attendant feature: the system administrator, registered users, and unregistered users. Each of these users can perform a specific set of SAA tasks. See Users for more information.
SAA can connect and interact with Microsoft Active Directory, the MiCollab, or with the NuPoint system mailboxes. For systems that are installed with NuPoint UM, the default data source is NuPoint UM, and all user mailboxes with FCOS bit 92 assigned are included in the data source. NuPoint UM administrators may also select Active Directory as the data source. For MiCollab systems, the data source is always MiCollab. For standalone systems, the default source is Active Directory.
The supported prompt languages for Speech Auto Attendant are North-American (NA) English, British (UK) English, Canadian French, and European French. The bilingual option allows you to combine any two of the four languages, setting one as Primary language and one as Secondary. Callers are prompted first in the Primary language and then in the Secondary. When properly configured, the Bilingual option allows callers who speak a language name (or an extension number) in the secondary language to be automatically prompted in that same language. Callers may also select a language using the dial pad.
Note: For MiCollab installations, supported languages are North-American (NA) English and British (UK) English only.
For system and department prompts, you can use the default greetings, import customized greetings, or you can record your own prompts.
SAA is an optional feature that must be licensed in order for the configuration fields to be enabled. SAA licensing includes the required number of telephony channels and the Text-to-Speech (TTS) core license that is required for all speech-enabled applications. (Note: The TTS core licensing is shared with the Advanced UM feature.) Licensing is performed on a per-user basis. When the maximum number of licensed users is reached, the administrator receives a warning message that indicates the number of users who are excluded. All licensing is done through the Mitel Applications Management Center (AMC). For more information about the AMC, see Planning and Licensing and the NuPoint UM Technician's Handbook. When you add additional licenses to the system, you must restart the server.
The speech-enabled channels are configured through Line Group assignment. A new type of Line Group is added to indicate that all the channels inside the line group are speech-enabled channels. The licensing of the total number of speech telephony channels is cross-checked when you configure the speech-enabled channels.
The diagram below illustrates the high-level system layout. The following table lists components and their relationships to SAA:
Speech Auto Attendant |
Outside Speech Auto Attendant |
||
Admin Web Console |
Speech-Recognition service |
Microsoft Active Directory |
Mitel Unified Communications Server |
Web View interface |
Update Server |
Microsoft LCS 2005 Server |
Mitel Application Suite |
VoiceXML Browser |
NuPoint database |
Microsoft OCS 2007 Server |
telephony network and users web browsers |
Text-to-Speech engine |
Grammar files |
|
|
Note: Microsoft OCS R2 (formerly Office Communications Server 2007 Release 2) is not supported for Release 4.0 of NuPoint UM.