Maple PDF-to-presentation workflow sample — checked 2026-10-12

maple-ppt-brief.pdf is a self-authored two-page source with fictional records.
The source records book sorting, not reading performance. A: 20 books, 8 minutes,
4 boxes. B: 30 books, 12 minutes, 6 boxes. Both aggregate rates are 2.5 books/min.

Two AI generation tasks used this same parsed PDF. The first completed but
could not open because of a formula syntax error; reopening returned the same
error. The second prompt requested plain-text calculations and produced four
editable-browser slides. This does not prove a universal error fix.

maple-ppt-original.pptx is the second task before manual edits. Its final
discussion question incorrectly mentions reading efficiency. The reviewed
version changes that question to sorting quality and replaces slide 4 notes.
All other slide and note parts and the embedded image are identical between
the corresponding native downloads. The core modification timestamp differs.

Both public PPTX copies replace one image-description attribute that contained
an internal resource path with an accurate figure description. No visible text,
notes, pictures or other uncompressed ZIP-member bytes are changed by this
publication cleanup. Original native downloads are retained separately.

maple-ppt-figure.jpg is the observed parsed-source image. It is byte-identical to
the JPEG embedded in both decks. The deck does not contain a native chart or
native table. Calculations are plain text. Four notes parts are present.

Screenshots on the site are Doc2X browser previews. PowerPoint/WPS playback and
editing were not tested. Confirm target-software fonts, layout and presenter
view before presenting. Same-browser reload restored the edited question and
note; cross-device sync and permanent retention were not tested.

This ZIP was assembled locally, not exported as a product package. See
checks.json and the two request files for the exact scope and checks.
