I have, btw, just finished reading “the phoenix project”. Brilliant book! Having lived through pretty much that story a couple of times, the solution is just that simple. Not easy, but simple.
Been around a while, at several big companies, I know most of them will do RPA (they are of course the ones having a blockchain project plodding along), but only a few will go down the “software company” route. It’s kind of tragic.
The problem comes when RPA stops being a short-term tactic, and starts being a long-term cottage industry. Unfortunately, RPA vendors have a vested interest in keeping A around as long as possible in order to maintain the necessity of their solution.
The RPA abstraction just seem so fragile and unflexible. I _had_ to build “rpa” like solutions to problems in the late 90s using tools like scriptit, and later vbscript. Crap software have no APIs...
I guess, it’s the same type of companies, still with the same type of software.
It will never completely disappear, but there is no future where a BA with an overhyped macro suite replaces the whole of software engineering.