How does a converse on step work.?? 1

[dimpoo]

How does a converse on step work.
Is this a proprietory step or will it work with a Nortel Periphonics IVR system.

 

[nohuhu]

dimpoo,

it will work with any ivr system. the technique is very simple: when the converse-on vector step is reached, definity places a call to the specified split/skill with specified priority. call will be distributed by ordinary acd rules (that you configured, that is). when a call is answered, definity will pass specified parameters (up to two) using dtmf tones and '#' character as an end-of-digits sign. if there are no parameters specified (none and none) there will be no tones at all. note that customer's call will be still on queue and when an agent becomes available call will be routed to that agent and the second call to ivr will be dropped.
for physical connectivity you can use either analog station ports with type "vru" or t1/e1 trunks with off-premises station type "ds1fd".

 

[dimpoo]

Thanks dwalin!!

That is great..

Just a question more.

I need to play the EWT announcement on the IVR.
I will converse on the IVR skill passing wait and none.

After this, will the IVR have to press the converse return code for the call to come back to the original position on the queue or as you have mentioned above, as this is a second call there is no question of it losing its position in the queue.

 

[nohuhu]

dimpoo,

no. converse return is used when there's a need to return some value from ivr back to the definity. in your case there is no such need because a call placed in queue to operators using queue-to step will not lose its position nor be affected in any other way except that caller will not hear music on hold from the moment the system passes all the digits to ivr. from that moment ivr is responsible for playing anything back into voice path otherwise a caller will hear silence. the call processing goes its normal way and when an agent becomes available call will be routed to this agent and at this very moment definity will drop ivr connection. unfortunately, there's no way to affect this to allow an ivr to finish an announcement or something. if you need an interaction with ivr that should not be disrupted, don't use converse-on step, place a call in queue with queue-to command, just like with ordinary group of live agents.
in your case, a vector should look like this:
01 queue-to split/skill x priority y
02 goto step 4 if expected-wait-time < zz
03 converse-on split ivr priority m passing wait and none
04 wait-time 10 seconds hearing music

in short, first place a call in queue to live agents, then check expected wait time to determine the need for converse-on step, then do converse-on. keep in mind that it is not advisable to have callers hear just a part of ivr announcement, better just to hold them in queue playing music. check your ivr announcement length, add 3 seconds needed for setting up ivr channel and passing digits and you will get zz number.

 

[dimpoo]

Thanks Dwalin,

I have actually setup a pilot, where i login on a Digital phone as a agent where the call would be routed incase i would converse on.

Upon picking up the call, i hear 2 tones, wanted to know in what format would the EWT be passed?

I know it is in seconds but, would i get something like a EWT space # or something?

Would be great if you can help.

 

[nohuhu]

dimpoo,

any information is passed in form of dtmf digits, followed by '#' sign. so in your case it may be one digit and #. ewt is passed in seconds.