Right, but Postman would have been a valid choice all the way up until it wasn't. The cost of building your own used to vastly outweigh the cost of having to possibly move to another app, pay for licensing when that changes, etc. But that's changing (especially for something like a GUI to make http requests which feels pretty mechanical for basic use cases).

Tangentially related, one belief I've had since the beginning is that LLMs will result in the decrease in large consulting/software factory companies (good riddance) and one-size-fits-all solutions in favor of solutions tailored to a business's actual needs. I work in a slice of healthcare. There are a handful of big EMRs and they were all built 20 years ago in an industry that changes and so many providers end up having to retrofit their businesses around whatever quirks/workflows the software has. And it's usually one of a few flavors: overly restrictive, too configurable, or worst of all: not restrictive at all. None of those are ideal. But now that the economics are changing, hiring a software dev or 2 at your small/medium-sized company becomes a lot more appalling. And it means they actually get business experience. Which means better software all around. I believe the days of shuffling changing requirements through 4 layers of PM's and BAs and sprint leads only to end up with a solution that looks like it was built playing the telephone game are waning.