How do we run parallel notification paths? Here is the use case:
When an escalation event is started the user will select a group for the recipients. This may be done manually by going through Messaging or via an inbound integration. That group (or groups) each has an escalation path for handling the event. The culmination is for whoever finally takes the event to mark it "fixed" and end the action.
However we also want a separate group to monitor the event. This group is charged with oversight to make sure the event progresses to "fixed" within a certain amount of time, depending on the severity of the event. So if the event does not get "fixed" then this group gets notified.
Since this "monitoring group" to be added to any event generated from the form. I thought just adding this new group to the form would solve the problem. But I don't think so, since adding a group to the form is for defaults and what we really want is "in addition to". Also changing the time bound based on the severity of the issue doesn't fit well into this.
Using Activity Updates or subscriptions (as they currently work) doesn't quite fit the bill. We want this team only to be notified if an issue is not fixed within a time period.
My current thoughts are a kludge. Create an outbound integration "monitoring group-start-trigger" pending on delivery of the form. Create a second outbound integration "monitoring-results". Create an inbound integration "start-monitoring-timer".
When monitoring-group-start-trigger gets the notification that the event has been delivered it uses the REST api to start-monitoring-timer that includes the URL of the event and its severity. Then start-monitoring-timer make ssure the new event and the original one are linked, checks the severity and sets the recipients to a monitoring group based on the severity. So for example severity-group-high would wait 2 hours and then send a notification to everyone in that group with a link back to the original event; severity-group-medium would wait 2 hrs and do that...
Finally, if someone does mark the original event as fixed, the monitoring-results outbound integration gets that, uses the link to the monitoring group event and terminates it, thus halting the timer.
I think this would work, but its complicated and kludgy. Is there a better way?