To be able to fully automate the FMEserver it is nescecarry to be able to create the Automations by API methods.
The idea is to extract all sources and targets of all workspaces or a set of workspaces on the server.
Put them in the correct order of dependency and try to parallelize as much as possible.
So the FMEserver is used most efficiently.
If yoou think about cloud and effincient use of Fme engines this is the most optimal configuration.