Create endpoint based on /stelle, grouping passages by texts and texts by authors, paging by authors

[kaggl]

There is a component (pictured), which shows a list of authors, their texts and the corresponding passages related to either keywords or case studies. Right now, this is done via doing a normal /stelle call with a very big limit and parsing this to get the desired result. This is problematic because of paging, ideally the paging is based on authors, so you can load another set of authors if you want.