What features of Pega 7 are lost when we make Pega as a Headless app?
You may be witnessing the new enterprise architecture trends / initiatives in building the UI in one layer, and building micro services with in the applications. This enables to adopt to the new UI standards without the dependency on the Product limitations. This initiative would be across the applications and these applications may be built using different tools and technologies. Not every technology offers the similar UI building capabilities as Pega 7 offers. So this initiative may work very well for the tools or technologies that do not have robust UI building capabilities. However, this approach with Pega will have an impact on the complete usage of the Pega 7 features.
I would like to understand from the Product Experts on the Pega 7 features that would be lost when we use Pega 7 as a headless app (building UI in a separate framework)?
If I understand your requirement correctly then you are planning to use your own UI component and want to implement only the core engine and other underlying feature of PEGA in your application. Am I correct ?
I think you then need to think of implementing only Business Rule Engine ( BRE) . So basically in addition to our BPM suite with its tightly integrated business rules capabilities, Pegasystems also offers an independent business rules engine that can power a higher order of business logic for your existing mission-critical business applications.And in that way you will get a chance to enjoy our world class BPM features except its UI components
I dont have any PoCs/experience to back this but my thoughts below
a) You might lose on functionality like Pega pulse etc which may or may not have formal public APIs associated with them.
b) Since micro-service communication is stateless, you will find security, authentication, authorization, session management a pain to implement.
c) You will probably be implementing UI logic (screen flows etc) in UI layer as opposed to Pega.
d) If you include UI as part of your micro-services stack (as opposed to leaving it to various channels to implement their own UI), you will probably have faster regression testing when you make changes to individual components.
Hopefully others chime in as well as I also need to know the answer to this question