For the case for parallel processes (the same data is spread across multiple connections).
For CCPM, you would write a Data Job for each source system and then merge all case, activity, and data tables.
Would you simply create multiple transformation scripts for each Object (one script per source system)?
Thanks