We have a client that has an existing NEC system that will let a caller call and go to an interactive prompt system. The system will ask the caller a question and they will give an answer and press #. There are around 6 to 8 questions and when it is done it takes all the answers and puts them together and sends it to a chosen voicemail. Is there any way for this system to do that? We have figured out how to get a script to run but we are not finding anything anywhere that we can save the file we create and then send it to an extensions voicemail. Thanks for the help and really have looked everywhere for this solution. Paying for a 3rd party integrated solution would also be fine.
Thanks dicko. I have been using the forums for quite some time but this is my first post. You are always pretty spot on. We aren’t looking to have a speech to text. We want a preloaded list of questions that we record and after every question the caller answers and presses # to go to the next question. Then when all the questions are done that recording of the answers is combined together and put in an extensions mailbox. Again thanks for the help. This one is a little stressful because it is a big feature for them. We are not afraid to program.
I am not understanding the logic here, if you only accept # to go to the next question, at what point will you be effective? Surely you need to differentiate at some critical point and so redirect, no?
@frankb Pretty much but in this case they are asking questions to help fill a prescription. After the last question is answered and they press # they exit and it says thanks for the information.
Either way, he needs to intercept the interaction with the client, either by voice (hard) or DTMF (easy) other than that the inbuilt IVR system should work just fine after parsing the voice/dtmf choice, The original post suggested a voice response was expected. Go with whatever works
@dicko I must be missing something but I see no way to do this with the built in IVR system. I have looked everywhere for a way. Again thanks for the help.
requires a voice to text solution, it could be “The Sound of Music” or maybe “Blazing Saddles”, you are back to cheap/expensive to differentiate and then a serious AI back end to even get close . Or perhaps I am missing some magic you expect?
We would create the recording for “What is your favorite movie?” We would like the user to speak the answer and the system record the audio. Save that audio and then after the next question append to that audio. When it is all done have it sent to voicemail.
The really new Asterisk also does that and always has, but not through the GUI, there is the record app you can use in a custom-app,
Record(filename.format[|silence][|maxduration][|option]) the options include appending
there are many examples out there.( perhaps from PIAF) that do that necessary precursor interactive bit, but you probably need a consultant. (150 bucks an hour would be fair)
You could do this using ARI. You would send the call to "stasis’, and then build a stasis app that receives the call event from asterisk. The coding of the stasis app can be built using any dev tool that you want.
In freepbx you need to do your recordings, and go behind the scenes to add some text to the extensions.conf (or rather the freepbx additional extensions.conf - can’t remember the name exactly). Let’s say you tell it to send extension 1000 to stasis. Create a miscellaneous destination of 1000 and then point your inbound calls to it.
The call comes in, gets pointed to stasis and waits for the stasis app to take over.
I am doing something similar at the moment (in C# though) but using Dtmf, but I know you can handle the recording side using ARI as well.
