LoreGraph product docs
Document processing turns uploaded source files into extracted text, sections, course outlines, lessons, and quiz drafts.
LoreGraph processes uploaded documents in stages: upload, text extraction, content cleanup, section detection, outline generation, lesson draft generation, and quiz generation.
PDFs can have page-count handling and may process differently based on size.
DOCX quality depends heavily on heading styles and clean document structure.
PPTX files may produce flatter outlines because slides do not always have a deep heading hierarchy.
Messy documents should be reviewed carefully before publishing generated content.
| Setting | What it does | Recommended default |
|---|---|---|
| Supported file types | Controls whether upload can proceed. | PDF, .docx, and .pptx. |
| OCR or scanned content | Affects extraction quality for image-heavy pages. | Use readable text PDFs when possible. |
| Outline review | Prevents bad structure from flowing into lessons. | Always review before generation. |
A slide deck with one process per slide can become a clear training outline. A deck with dense speaker notes and no slide titles may need manual outline cleanup.
Last updated: May 30, 2026