[GenAI] Generate personalized feedback for answers to open-ended questions

As a proofreader, I can validate and rework automatically generated personalized feedback against expectations: an exemplary response and/or correction criteria.

It's energy-consuming and tedious to analyze and write feedback from scratch

How to avoid bias in corrections & analyzing each response

.