With all those rapid experiences in teaching maths online, I see many occasions where the wheel is reinvented. For example, because popular English platforms mostly just allow one answer, platforms that allow reasoning steps seem novel. But they've been around for decades...

For example end 80s to around 2000 Beeson's Mathxpert

Aplusix French think around 2005. Still around, web.

Own PhD research used online applets equation solving. Think 2002 to now. Used in many Dutch maths textbooks. Try http://is.gd/kes2017  as guest for many equation apps (and others e.g. famous block building app).

Around 5 years ago already some threads with @timstirrup about Mathspace.

I guess the time needs to be right for some innovations. But in doing so I wish there was some more awareness of of this. By building more on prior work, invest and collaborate more platforms could already have been much better than multiple-choice and/or one answer.