I've been working on a workaround for this. I've created individual connections for each CSV file. At a certain point in my package, and while using precedence constraints for each table, I have been able to connect to each file individually for my ETL processes. the attachment sort of shows.