They're used interchangeably but they probably shouldn't be. What characterizes engineering (in most disciplines) is the practice of modelling. An engineer know

They're used interchangeably but they probably shouldn't be. What characterizes engineering (in most disciplines) is the practice of modelling. An engineer knows exactly what they're going to build and knows exactly what characteristics it's going to have. At least, they know the ones that are knowable, and they're aware of the ones that aren't. Do you evaluate solutions quantitatively, like "This solution will require n times more memory, and it'll run m times faster". Or "this is how many nodes we need to guarantee a typical latency of < x milliseconds assuming y active users per minute" Or do you say "let's try this and see what happens"? The latter approach is very common in the software industry, but (thankfully) not in bridge-building or space exploration. So there ought to be terminology that distinguishes these practices from each other.