On Fri, Mar 11, 2011 at 1:36 PM, Robert Harrison <[email protected]> wrote: > >> There's the rub. We can't send the police a text and they don't monitor >> email. It has to be a voice message. It could even be a canned message. > > The concept is very cool, but whatever your message is how are you going to > tell when they actually answer? If they put you on hold how will the system > know? I suggest putting your message in a never ending loop, but if they > answer on a part of the message that sound like a recording they might just > hang up. > > How do you plan to deal with that part of the problem?
Determining when a call is answered is pretty straight forward as it generates a SIP event and whether you got a human or an answering machine/voicemail is done through a process called CPA (Call Progress Analysis). Getting put on hold is more complicated scenario. I'd say that the best way to deal with that is to have a loop in your voice xml dialog that waits for user feed back in terms of either listening for a spoken phrase, "Say 'OK' when this message has been successfully delivered" or a dtmf tone "Press 1 to confirm that the message has been delivered". I'd also make sure that CPA detection happened continuously during the call and have a new CPA event kick off a restart of the dialog. That way if the call gets put on hold, it will be looping in the background, but when it gets pulled out of the hold queue and someone says, "Hello?" it will trigger a Human CPA event and restart your dialog that plays the "we really need help" message. Judah ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Order the Adobe Coldfusion Anthology now! http://www.amazon.com/Adobe-Coldfusion-Anthology/dp/1430272155/?tag=houseoffusion Archive: http://www.houseoffusion.com/groups/cf-talk/message.cfm/messageid:342940 Subscription: http://www.houseoffusion.com/groups/cf-talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/groups/cf-talk/unsubscribe.cfm
