Hi, What would adding execution and analytics engines offer us in te

The best way to come up with the right number of Execution/Analytics engines is contacting your account executive to enroll in Appian Labs. The analysis of your current usage (among other things) plus the expected growth can determine the needs of your architecture; it's not something that can be determined without a proper and deep analysis.

Just to answer your main question; process execution engines are in charge of process instances while analytics is in charge of reporting on those instances, therefore the addition of execution engines gives performance improvement, for example, in environments where the data is still process centric rather than database centric.

I don't have any firm numbers to support this, but we recently added more engines to our dev/test/prod environments (combined with building a 2nd server for each and putting some old and some new engines on each).

This combined with some other server tweaks appears to have helped our server responsiveness. I know the extra engines use more RAM, but the majority of it is process based anyway and the extra few hundred MB per engine is nothing compared to our rather heave usage for processes.