This is intended as an intermediate "bridge" step that will let us take the output of GPT3/4 and use it to author the skeleton of an interview in JSON format. The Weaver uses the skeleton + the original file (DOCX or PDF) to make a full-fledged Weaver interview.
This is a new, experimental feature that I might demo at Jurix next week, and it doesn't interfere with existing features. I've tested it. Going to merge for now but I welcome more refinements.
Fix #905 Create an API for #907 (not used yet)
This is intended as an intermediate "bridge" step that will let us take the output of GPT3/4 and use it to author the skeleton of an interview in JSON format. The Weaver uses the skeleton + the original file (DOCX or PDF) to make a full-fledged Weaver interview.
This is a new, experimental feature that I might demo at Jurix next week, and it doesn't interfere with existing features. I've tested it. Going to merge for now but I welcome more refinements.