These kinds of products are extremely hard to get right.
You likely end up in an awkward position where big or serious companies don't take you seriously because they have their own ways and methods and teams and baggage and don't want someone else's magic touch.
And on the other end of the spectrum you have noobs who need way more hand holding than your stuff provides so your product is too technical for them.
Another problem is a case of "death by a thousand cuts". You inherently end up dealing with an infinite combinations of environments and requirements and you can't fully satisfy any one group of customers because each one is just different enough to cause a problem.
So put simply I would say something like this is too abstract and too general for a small operation.
Perhaps companies like Amazon can afford to dump resources into projects of this nature because they have the resources and they have the eco-system to put it into. But I would recommend against trying to bootstrap projects of this nature.
As a generic advice I would suggest attempting more concrete problems that have bright and clear boundaries and a finite and small domain so that you have a very clear "finish" point (and you will still never be finished).
A completely made up example to demonstrate generic vs specific, instead of "DevOps as a service" think "Hosted Redis as a service".