Field notes from the PDF factory.
AI Redact v2: 10× faster, now with custom patterns
Our detection model just got a rewrite. Here's what's new, how we trained it, and why batch jobs now run in seconds.
A practical guide to Bring Your Own Key
When BYOK makes sense, when it doesn't, and how to wire up Anthropic + OpenAI in under two minutes.
PDF security in 2026: what every team should know
Encryption, redaction, metadata leaks, and the quiet problem of OCR residue.
Five AI workflows legal teams actually ship
From clause extraction to conflict checks — the playbook we hear most often from AmLaw 200 firms.
Announcing the pdfcraft ai API
REST, webhooks, SDKs, and a free tier for hobby projects. Build with PDFs like you build with Stripe.
Why our summarizer cites page numbers (and yours should too)
Hallucinations are a trust problem. Here's the technique we use to make every claim traceable.
How to merge PDFs without losing bookmarks or hyperlinks
The default merge in most tools strips bookmarks. Here's how to keep them — and reconcile internal links to the new page numbers.
Compress PDF for email — the 5 MB problem
Gmail caps you at 25 MB. Some corporate gateways cap you at 5 MB. Here's how to land just under any cap without making the file look like a fax.
PDF to Word conversion: when it works, when it doesn't
Some PDF-to-Word conversions come out clean. Some come out as a wall of text-frames. The difference is what was in the source.
How to OCR a scanned PDF and make it searchable
Cmd+F doesn't work on scans because there's no text layer. Here's how to add one without changing how the document looks.
Splitting PDFs by range, size, or bookmark — which mode to pick
Splitting page-by-page is different from splitting by file size, which is different from splitting at section bookmarks. Each fits a different job.
How to translate a PDF without breaking its layout
Plain-text translation gives you a wall of text. Layout-preserving translation gives you a usable document. Here's the difference.
Chat with PDF: 5 prompts that actually work
Vague questions get vague answers. Specific, locatable questions get precise answers with page citations.
How to redact PDFs properly (and why most 'redactions' fail)
A black rectangle on top of text isn't redaction. Real redaction permanently removes the bytes.
Sign PDFs free: typed, drawn, or uploaded — what's legal?
Visual signatures vs cryptographic signatures. Most business contracts only need the first.
How to extract tables from financial PDFs into Excel
10-Ks, balance sheets, income statements — the data trapped inside PDFs into a spreadsheet you can chart.
Edit a PDF in your browser: what's possible without Acrobat
Most PDF edits don't actually require Acrobat. Here's what works in the browser, and what doesn't.
How to compare two contract versions without Word
Counterparty sent V2 back. What changed? Here's a 30-second redline you can read like a lawyer would.
Add page numbers, watermarks, and headers in 30 seconds
Three operations that look like three tools but are actually one.
Convert images to PDF the right way (phone photos, scans, etc.)
JPG-to-PDF looks easy but the choices matter: page size, fit mode, OCR, ordering.
How to summarize a 200-page report in plain English
Eleven different summary formats, each calibrated for a different audience.
Why your scanned PDFs need OCR before anything else
OCR is the prerequisite for almost every other operation on a scanned PDF.
How to combine bank statements for your accountant
12 monthly statements → one searchable PDF in 5 minutes.
Compress vs Optimize: what each level actually changes
Light, Balanced, Strong — three levels with different trade-offs.
How to fill out a non-fillable PDF form
Old forms scanned without fillable fields. Type in them anyway.
The 7 PDF mistakes that cost businesses time
From 'redact with a black box' to 'compress signed PDFs' — the recurring patterns that waste hours.