Using Zammad to streamline communications would be even more powerful if we can deal better with external systems that send information to Zammad without having a proper integration to be context aware. In difference to human input (that mostly catches up with the context via replying to Zammads notifications) external systems without integrations have usually structured syntax and data that could be parsed and processed.
This would cover widespread external systems like JIRA, Zabbix, GitLab & co when they send mail notifications in repetitive form and would act as the lowest form of incoming integration. It would especially be useful for streamline communications in B2B support scenarios where also the enterprise customer communicates instead of manually strictly via support systems that either have no Zammad integrations or won’t get sufficient access for web hook integrations.
Provide some kind of profile editor for Zammad administrators where they could in self service define certain pattern matching via regex on a limited scope incoming mail. With the postmaster filters we could presort all communication from an external system (i.e. firstname.lastname@example.org) for then run pattern matching on them and try to run some magic on those limited scopes. The work flow could be to first run the channel filtering for not breaking stuff outside scope, then parse variables with a regex and run the actually processing like ticket or article creation.
Of course all hipster machine learning could do here all fancy magic but this is probably not needed for simple cases like i.e. for all messages from an external legacy help desk system extract the issue ID, search in open Zammad tickets from the same external system for that ID and add the message as an article in the existing ticket. Next level are triggers on i.e. messages from Zabbix that have beside an incident ID and a state information to act upon with the state of the Zammad ticket.