Implement first complete workflow

Describe the feature you'd like Implement a first complete workflow for creating real benchmarks from real data. This encompasses the following steps:

[x] get the data from quiver-data repository
[x] convert a ocrd process workflow to a NextFlow workflow with the OtoN converter
[x] execute this workflow on the existing data (incl. evaluation)
[x] create a result JSON according to the specs
[x] push this JSON to the quiver-back-end repo

Additional context

Issue for the corpora

OCR-D / zenhub

Implement first complete workflow #131