Designing pipeline on the fly

Hi Gurus,

My requirement :
Create snaps for multiple source ( Oracle,flatfile,iseries) or targets (APIs/Flatfiles) or mapping rules ( like source row1 goes to row4 in target) and
on the fly create and run a pipeline using these snaps via some GUI (.Net/Python)

Basically I want o create and execute an pipeline from GUI on the fly… Is it something doable ?

So if a user select Source A ( Oracle for example), within source A Table B ( Employee table for example) and witin table b only certain number of columns (column 3-7 for example) to ftp in some ftp location by reordering columns from table b via GUI ( some .Net framework), Snaplogic job should read that from GUI and run and drop that file to ftp location mentioned in GUI.

1 way is to create n number of pipeline ( 1 for each of the usecases) and run every one of them based on UI.However this will restrict as any new source would have to be redesign.