Community

Kanan, my apologies for not responding sooner. First, I would like to explain what I mean BPM before I address your question.  To me, BPM is about the automation of processes and/or human interactions as it spans multiple applications; if you like, the “inter-application process logic”. When done well, it will use services exposed by the backend to abstract the details of the process and keep it resilient to change. Likewise, at the front-end, BPM should only involve a user when the process paths are easily modelled and known ahead of time. To the extent that these services and rules are in effect standards, then you are right, having these defined will result in more robust solution.

The problem with many RPA solutions is they operate as a synthetic end-user and hence they either cannot or do-not directly leverage the back-end services. This means that changes to the UI (which are largely ungoverned) have the potential to break the ‘robot’. Furthermore, RPA’s promise of not requiring any changes to the applications makes it impossible to augment the user journey to allow additional steps for intervention and review. Sure, the lack of standards here is an issue – but the bigger problem is trying to use RPA for exploratory processes. Once an organisation has exhausted the pool or fixed/repeatable processes, then another approach is required.

James