Not really interested in naming and shaming because the lead dev trolls the internet to find people that don't like the framework it so he can publicly mock them.

As for PyPika, I wouldn't be fond of that, but that at least has an ability to get the raw SQL being built. There's no such option in my framework (and I've looked for hours combing the source code trying to find it), so debugging why the data coming out is wrong is kind of a nightmare (especially since the documentation is incoherent).