SQL: Calculate duration based on dates and parameters (Change Log) - sql

Question

I have a dataset that is like a ticketing system change log, I am trying to calculate kind of like an SLA time across the records, like how long did this specific ticket sit with this Group 2 for before it was either resolved or moved to another group to resolve.
The data looks like so:
ID
field
value
start
end
1
assignment_group
Group 1
2022-03-21 08:00:00
2022-03-21 08:05:00
1
incident_state
Work in Progress
2022-03-21 08:05:00
2022-03-21 08:30:00
1
assignment_group
Group 2
2022-03-21 08:35:00
2022-03-21 08:50:00
1
assigned_to
User 1
2022-03-21 08:50:00
2022-03-21 08:51:00
1
incident_state
Work in Progress
2022-03-21 09:00:00
2022-03-21 09:30:00
1
incident_state
Resolved
2022-03-21 09:30:00
2022-03-21 09:31:00
2
assignment_group
Group 2
2022-01-21 11:30:00
2022-01-21 11:35:00
2
assigned_to
User 1
2022-01-21 11:35:00
2022-01-21 11:37:00
2
incident_state
Work in Progress
2022-01-21 11:40:00
2022-01-21 11:55:00
2
assignment_group
Group 3
2022-01-21 11:58:00
2022-01-21 12:00:00
2
assigned_to
User 2
2022-01-21 12:05:00
2022-01-21 12:06:00
2
incident_state
Resolved
2022-01-21 12:10:00
2022-01-21 12:07:00
The issue I am having is calculating the duration based on the start time the ticket was assigned to a specific group and the end time of when either the ticket was resolved by that group or moved to another group to resolve. For example, I am only interested in Group 2, the duration for ticket 1 for when the ticket was sitting with Group 2 till the ticket was resolved by Group 2 is 2022-03-21 08:35:00 to 2022-03-21 09:31:00, so duration is 1 hour and 1 minute. But for Ticket 2, the ticket sat with Group 2 from 2022-01-21 11:30:00 till it was transferred to another group to resolve at 2022-01-21 11:58:00.
My code looks like so at the moment, I join two tables to pull in the ticket information and then the ticket state changes (so every time an action is taken on that ticket). Then I am left with the table above. I kind of guess I need to use a lead function but I can't figure out how to get the correct end time for the correct record (When incident_state = resolved OR assignment_group = another group):
WITH incidents as
(
SELECT number, sys_id AS SYS_ID_INCIDENT
FROM tables.ServiceIncidents
),
changes as
(
SELECT id, start, field, field_value, value, `end`
FROM tables.IncidentInstances
),
incident_changes as
(
SELECT *, TIMESTAMP_DIFF(changes.`end`,changes.start, MINUTE) as Duration, row_number() over (partition by number order by start) as RN
FROM incidents
LEFT JOIN changes
ON (incidents.SYS_ID_INCIDENT = changes.id)
),
IAMtickets as
(
SELECT i.number, i.SYS_ID_INCIDENT, start, field, value, `end`, Duration, RN
FROM incident_changes i
INNER JOIN
(SELECT DISTINCT number FROM incident_changes WHERE value ='Group 2') r
ON i.number = r.number
),
cte as
(
SELECT *, lead(value) over (partition by IAMtickets.number order by start)
FROM IAMtickets
)
SELECT * FROM CTE
I want the output to be something like this:
ID
Assigned to
Duration
Outcome
1
User 1
1 hour 1 minute
Resolved
2
User 1
28 minutes
Transferred