Summarizing Timestamp data from multiple rows

[mosterman]

Good afternoon,

I've spent several days trying to find a solution but having had no luck... 

I have a dataset of phone records from a contact center showing every change in an agent's phone state - when agents take calls, are available, are at break or lunch or meeting, etc.  Each event is generated as a line item with the following information: 

inqueue	Agent ID	Start Time	End Time	Description
Was the agent on a call or able to take a call?	The agent's unique ID from the phone system	The start date and time of the event	The end date and time of the event	The event's name/description

To give an example: 

inqueue	Agent ID	Start Time	End Time	description
TRUE	15	2/22/2023 9:52	2/22/2023 9:52	Available
TRUE	15	2/22/2023 9:52	2/22/2023 9:55	ACD
TRUE	15	2/22/2023 9:55	2/22/2023 9:56	ACW
TRUE	15	2/22/2023 9:56	2/22/2023 10:02	ACD
TRUE	15	2/22/2023 10:02	2/22/2023 10:03	ACW
FALSE	15	2/22/2023 10:03	2/22/2023 10:16	Break
TRUE	15	2/22/2023 10:16	2/22/2023 10:18	Available
TRUE	15	2/22/2023 10:18	2/22/2023 10:19	ACD

So in this example, from 2/22, the agent was in queue from 9:52 AM to 10:03 AM.  Then they were out of the queue from 10:03 AM to 10:16 AM. They were then in queue from 10:18 AM to 10:19.  (Technically, it would be longer than that but I didn't want to throw dozens of rows into this)

Using the data above as an example, my desired output here would look like this: 

inqueue	Agent ID	Start Time	End Time	description
TRUE	15	2/22/2023 9:52	2/22/2023 10:03
FALSE	15	2/22/2023 10:03	2/22/2023 10:16	Break
TRUE	15	2/22/2023 10:16	2/22/2023 10:19

In short, giving me a start and end time for each sequence of events where the inqueue value is the same.  (So if I have 25 consecutive records where inqueue is TRUE, I should get a summary of the start time and end time for just that sequence.)

I have looked at the Summarize tool - but could not figure out how to get the distinction of these breaks in the sequence. I've looked at multi-row formula, but the problem is that not all sequences are the same number of rows.  If I were using Excel for this I'd have a few tricks I could use... but I need this in Alteryx because there is a lot more data than Excel can handle. 

Any help is appreciated!  Thank you in advance! :)

[SPetrie]

Something like this may work for you. I use a multi-row to just numerate when inqueue changes from true to false as its SequenceID. Then use it as one of the groupby conditions for the summarize. I also converted the date/times to date/time values instead of string to make sure that the min/max values are accurate.

I also used a formula to capture the descriptions of the false items and group by it as well.