Hi Product Hunt 👋 I'm Kevin, the maker of RifflePages.

PDFs on an e-reader are painful: tiny text, zooming, scrolling. Scanned books are worse. Most converters keep the page pictures, or copy the scan's old, broken text layer.

RifflePages rebuilds the book instead:

• reads every scanned page again with OCR

• finds the chapters and builds the table of contents

• removes running heads and page numbers, and joins paragraphs split across pages

• keeps italics, footnotes and illustrations

I tested it on public-domain scans from the Internet Archive: A Christmas Carol (1915, the one in the video), Frankenstein, Moby-Dick, Walden, Alice and Sherlock Holmes. Each one taught the engine something new, like chapters starting mid-page, or OCR reading "Adventure II" as "Adventure 1W".

How it works: upload a PDF, no account. You get a free preview (cover + the first 8 pages) to check on your e-reader. If it looks right, the full EPUB is US$1.99.

🎁 For Product Hunt: code HUNT-LAUNCH makes the full book free for the first 30.

I'd love to hear where it fails. If a book comes out wrong, use "Report a problem" on the result page and I'll fix the engine.