Lossless PDF cleanup
Automatically detects NotebookLM and other text/image watermarks and rewrites the PDF content stream instead of painting over artifacts.
About
About Deck Cleaner
Deck Cleaner focuses on NotebookLM slide decks and lightweight PDFs. The front end runs on Next.js App Router while the backend talks directly to the DeckCleaner synchronous clean endpoint to obtain a download token as soon as the document is processed.
The Charter and Solution specs define the guardrails we follow: ≤50MB / ≤200-page files, sub-10-second median processing, and UI surfacing of progress plus cleaning stats so users can confirm results quickly. Future releases will expand queue-driven jobs, feedback loops, and preview generation.
One-click browser workflow
No desktop installs. Upload → clean → download within the browser while DeckCleaner handles the heavy lifting.
Privacy first
Inputs live in temporary storage with minute-level expiry so sensitive documents never persist after download.
Want to dive deeper into the architecture or API? Check specs/charter.md, specs/solution.md, andspecs/api/openapi.yaml. We iterate on these specs frequently to keep Deck Cleaner fast, safe, and predictable.