callix.

Tools for difficult documents.

Valuable data is locked up inside complex documents that are optimized for human readability, not automatic extraction. These tools came out of solving that problem.

01Live

docscrubber

Strip names, addresses and account numbers from PDFs, right in your browser.

Open a PDF and docscrubber finds the personal data hiding in it — names, addresses, account numbers, phone numbers, emails — then takes it out for good. Nothing is uploaded, because nothing needs to be.

Local-first
Your document stays on your machine. No server, no upload, no copy of your file anywhere but your own.
True deletion
Redacted text is removed from the PDF itself, not hidden under a black box that anyone can drag away.
Review before you save
Every detection is listed, so you can uncheck the false positives before the file is written.
91.2% recall
Measured on real documents with pdf-pii-bench. Tuned to catch personal data rather than to look tidy.
docscrubber.appFree — no credit card
02Waitlist

quantable

Turn the tables in research papers into data you can actually use.

Draw a box around a table in a paper and quantable hands back the grid — headers, merged cells, units and error bars intact. Check it in an editable preview, then export to CSV or paste it straight into a spreadsheet.

Built for messy input
Skewed scans, low-resolution copies and watermarked pages, not just clean digital exports.
Borderless tables
The LaTeX-style tables that arXiv preprints and journal PDFs are full of.
Structure preserved
Nested headers, merged cells and multi-level groupings survive the trip.
Notation intact
± errors, footnote markers, significance stars, units and Greek symbols come through as written.
getquantable.comJoin the waitlist

Got a document problem that isn't on this list?