I am currently assessing client processes where the System of Record is a set of Mainframe systems and they are performing the activities directly on mainframe for each case. The client has implemented some kind of integration between Pega and one of the mainframes, but are hesitant to proceed with any further integration with other mainframes, or even with the other modules within the same Mainframe which was integrated earlier.

In such a scenario, I want to understand if RPA like Openspan would be an quicker option to implement compared to the BPM solution considering that RPA doesn't involve any back-end integration between Pega and Mainframe databases. Also, kindly share the process/steps to integrate Pega with a Mainframe system, and difference in cost between the two approach,

1. Pega integration with Mainframe

2. RPA for the process steps and calling the RPA from Pega once case is received