NOTE: This article has also been published in the official Exchange 2007 documentation. We recommend that you check the documentation for the most up-to-date version. Please see the following links:...
Updated Jul 01, 2019
Version 2.0
Blog Post
Thanks for the pointer Ross. I took a look at that page, and its very good for SIP provisioning (60-100 concurrent users) and erlang calculations. Does anyone know how this is affected by dtmf versus voice commands? Speech recognition is much more CPU intensive then simple dtmf processing, and I'm curious to know if the 60-100 concurrent users on a single box would stay the same if everyone was speech activated. It also says no accounting was done for automated attendants. I'd hate to deploy a UM solution and have to create a second UM server/dialing plan just because I didn't have voice vs dtmf sized properly.
Another open question is that since Ex2k7 now has inherent speech server functionality, has anyone seen accuracy numbers on corporate or personal contact dialing (the rate it recognizes the name properly). Part of sizing, since recognition is hurt by straining system resources.
Another open question is that since Ex2k7 now has inherent speech server functionality, has anyone seen accuracy numbers on corporate or personal contact dialing (the rate it recognizes the name properly). Part of sizing, since recognition is hurt by straining system resources.