Probably I didn't get the following idea across properly: this is not searching a company dataset. This is just like Google - as in it searches web pages, but with the added ability to filter by owning company industry, size, location, or later once we support it even stuff like tech stack used etc (in that regard it's somewhat adjacent to tools like Wappalyzer), or the HTTP headers present, or any kind of crazy stuff.
If that's still not unique enough (not sure TBH) I guess I should adjust the copy :)). I'm not great at writing copy.
Anyway this is still no a fully fleshed out idea, mostly just scratching my own itch and conquering the difficulties of storing billions of pages.
Thanks for the question and any feedback is appreciated.