autonomous ai managers will always be hard, because every company is different, thus you can't easily re-sell it as an enterprise solution. above I described t

autonomous ai managers will always be hard, because every company is different, thus you can't easily re-sell it as an enterprise solution. above I described the case where you want to monetize it, but I guess you're only interested to implement this in your own company? there's no reason it won't work. think from the first principles: as long as you have the time and resources for it, with code you can practically create anything, these things could even be done before ai, because you create an infallible process with cheap-human-in-the-loop-that-no-matter-how-stupid-cannot-mess-up, lots of great companies have incredibly fool-proof processes so they can pay cheap hourly wages. my take is yes, there's no reason this won't work if it's for a single company, personalized for some specific processes with some well-defined feedback loops. will it be costly/annoying? of course. will the ROI be worth it? I can't know without seeing your specifics