I have a project requirement to split recordings during a call based on the information being discussed in the call. In addition, these separate recordings need to be stored on different servers based on that content. Specifically, we will soon be handling Dept of Education callers discussing Federal Loans on the borrower account. However, many of these callers may have private loans with us as well. Dept of Ed has the right to "live monitor" callers discussing the Direct Loan data, but not the ISL private loan data. I've written a handler which currently sets a call attribute for the call type, either set to ISL or DL. When the client button is pressed (which calls this handler), the handler determines if any DL representatives are listening in on the call, grabs their CallID from the Monitoring session, and kills it. Currently is also issues a STOP and START from the record call toolstep, but this seems to only create segments on the call. How can I completely break the recording during this event, and automatically start a new recording prior to the agent discussing the other type of loans? These separate recordings need filed in a different hierachy on the recording server as well (hopefully based on the Call Attribute). The caller will NOT be transferred (better customer service this way). Any assistance is GREATLY APPRECIATED!!! Thanks all!

As the CallID remains the same and any attribute you set in the call will be effective for the entire call (this closes the possibility to categorize the calls segments in a different manner and then copy them to a different storage location), it seems unlikely to me that what you describe can be accomplished. Anyway, you'll better check with support or PSO.

Can the CallID be copied to a new CallID, then kill the original while allowing the new CallID to create a new recording, etc? A conference tool would work, but obviously not be preferred here. The main criteria are: * Can not transfer caller to other agent or workgroup * Needs to be automated solution, invisible to agents and caller * Must break the recording for original call and file separate from proceeding portion of the call still in progress. * May be repeatable during the new call (caller again wants to discuss other types of loans, ie. ISL --> DL and back to ISL again)

I just wanted to provide some details of our chosen solution. The process actually turned out to be pretty simple. In our custom transfer handler, I added logic to check a call attribute. This call attribute gets set to the type of call in progress. In other words, if the caller dialed our PRIVATE loan numbers, it is a private call. If they dial the FEDERAL lines, it is a federal call. Transfers look at the new destination (user, queue, station, etc) and grab a custom attribute which will be either federal or private. If the original termination point, and the new termination point are the same, the call is transferred. If they differ, we know we are crossing our boundaries, and the handler uses the PRIVATE tool step (not be be confused with the PRIVATE call type) make the call private. This breaks all monitoring sessions and recordings. The call is then marked as non-private (so a new recording can begin). As the caller is tranferred to the new destination workgroup, a new recording is initiated. This also automatically allows for storage of the separate recordings in different physical locations. In the recorder Categories container in IA, we have set a different compressed recordings directory path for the various types of calls. So far this looks very promising, and took very little coding. Gotta love I3 and their provided tools! :-)

That's a very innovative solution, which we love to see! Thanks for sharing!

Originally posted by bwilkening;23164
I just wanted to provide some details of our chosen solution. The process actually turned out to be pretty simple. In our custom transfer handler, I added logic to check a call attribute. This call attribute gets set to the type of call in progress. In other words, if the caller dialed our PRIVATE loan numbers, it is a private call. If they dial the FEDERAL lines, it is a federal call. Transfers look at the new destination (user, queue, station, etc) and grab a custom attribute which will be either federal or private. If the original termination point, and the new termination point are the same, the call is transferred. If they differ, we know we are crossing our boundaries, and the handler uses the PRIVATE tool step (not be be confused with the PRIVATE call type) make the call private. This breaks all monitoring sessions and recordings. The call is then marked as non-private (so a new recording can begin). As the caller is tranferred to the new destination workgroup, a new recording is initiated. This also automatically allows for storage of the separate recordings in different physical locations. In the recorder Categories container in IA, we have set a different compressed recordings directory path for the various types of calls. So far this looks very promising, and took very little coding. Gotta love I3 and their provided tools! :-)

I agree. That is a nice trick! Anyway, it looks to me that some transfer on the call is performed here, which I thought (per your previous post) shouldn't happen. Can you confirm if you finally allow the call to be transferred based on the call subject?

The issue concerning transfers is related to call recordings. We needed a way to segregate the call recordings for our PRIVATE loan callers (we own the loans) and the Federal loan callers (where the Federal Gov't usually owns the loans). Due to changes in the law, the Dept of Ed has the right to monitor callers for Federal loans, but we don't want them to listen in on the private loan portion of the calls. Callers may have both Federal loans (which we service for the Dept of Education), and Private loans (many branded types from various providers). As they choose to speak about one type or another, we needed to be sure the monitoring sessions were broken when they cross these "boundaries". Workgroups are isolated, and agents (while able to participate in FED or PRIVATE queue's) will be active in only one type of region at a time. They may be active in many Fed workgroups (dialer, inbound/outbound, etc), or many Private, but not both. I've written handlers to check the active workgroups and alert if any agent receiving a call is active in both Fed and Private simultaneously. I send automated emails to all Supervisors from all workgroup which they are members of, so the issue is immediately addressed (should a supervisor leave them active in multiple workgroup regions when moving them around to meet call demands). So, in summary, it is OK to transfer callers as long as anytime they cross boundaries we break monitoring sessions, terminate call recordings, and separately file each recording in physically diverse locations on our servers. I believe I have accomplished this, though massive full-scale testing is yet to occur. Thanks for the feedback. If you have additional questions, or comments for my consideration on how to make the process even better, please share. I'm always up for new innovative ways to use the I3 development tools!