PureEngage On-Premises

Hi,  We are a recent on premise genesys implementation with both GCXI and Pulse.      When comparing the data between the two system, we are able to get our reports to match with 99.5% accuracy.  For example, our calls offered for one day are off by about 50 calls.  (Out of around ~10,000) calls.

I'm looking for a 'simple' explanation to provide to my business user that explains this difference.   Any advice.
#Reporting/Analytics

------------------------------
Dawn Maciejewski
NiSource
------------------------------

That is a very accurate comparison.  Are you a 24x7 shop?  If so, it could be rollup time stamps that is causes them vary.  Where calls are getting rolled up to the previous day or following day based upon the event's timestamp.

It could also be some events between components are being dropped and not getting to ICON or StatServer, etc. 10,000 calls x __ events per call.. can lead up to 100,000k messages traversing the network.  As such a few network messages being lost is possible.

Unfortunately, to really know, you'd have to analyze the logs of ICON and StatServer and trace every call id to see if all messages appear or not.


Hopefully someone has a better answer.

------------------------------
Tony Morrow
Sr. Genesys Engineer
------------------------------

Just some thoughts.

GCXI is a historic  solution based on events from Icon.

While Pulse is real time solution getting its metrics from stat server.

Can you look at how your metrics are defined.  There could be some call flows that are counted differently.

It would make it easier to trace these sorts of issues if you could reduce the time frame to an hour or half hour period.



------------------------------
Andrew Soroka
Genesys - Employees
------------------------------

Original Message:
Sent: 04-07-2022 11:59
From: Tony Morrow
Subject: GCXI versus Pulse

That is a very accurate comparison.  Are you a 24x7 shop?  If so, it could be rollup time stamps that is causes them vary.  Where calls are getting rolled up to the previous day or following day based upon the event's timestamp.

It could also be some events between components are being dropped and not getting to ICON or StatServer, etc. 10,000 calls x __ events per call.. can lead up to 100,000k messages traversing the network.  As such a few network messages being lost is possible.

Unfortunately, to really know, you'd have to analyze the logs of ICON and StatServer and trace every call id to see if all messages appear or not.


Hopefully someone has a better answer.

------------------------------
Tony Morrow
Sr. Genesys Engineer
------------------------------