The engine behind this site was built for FDA label surveillance: comparing prescribing information revisions and reporting every changed warning, dose and contraindication. It applies the same rigour to SOPs, protocols and submissions.
Product labels, SOPs, protocols, regulatory submissions, CMC documents.
Text is extracted from both PDFs with its position and style. Repeating headers, footers and page numbers are removed, unchanged tables are recognised and held out, and paragraphs are aligned across the two documents before a character-level comparison. The output is a change list with page numbers, a redlined PDF and a side-by-side view with synced scrolling. The engine is deterministic: the same two documents always produce the same result.
Use the web tool for individual reviews, the REST API to trigger comparisons from your document system, or a private deployment in your own VPC or on-premises when documents cannot leave your network.
Yes. The diff is computed by a rules-based engine, not a language model, so the same two documents always produce the same result. That matters for validation.
Enterprise deployments run inside your own infrastructure with pinned versions, which supports IQ/OQ documentation.
Bring two real revisions to a 20-minute call and watch the comparison run.