If you have a backlog of paper or PDF contracts, the AI importer turns them into searchable, audit-trail-ready records in OMB Cloud.
How to import
- Open
/app/contracts. - Drag-drop one or more PDF/PNG/JPG files onto the import dropzone (max 10 files, 20MB each).
- Each file gets a status row: queued → extracting → done / error.
What gets extracted
The AI agent reads the document and pulls: contract title, parties (provider + client), signing dates, total value and currency, term length, list of signers, and the full contract body transcribed verbatim with HTML structure (headings, paragraphs, lists, signatures).
Image vs scanned PDF
Both work. Scanned PDFs (image-only, no text layer) take an extra few seconds. Unreadable sections are marked [illegible] rather than guessed — OMB Cloud never invents content.
Word documents
Export your Word file to PDF first. The AI extractor doesn't parse .doc/.docx directly. PDF is universal.
What you get
Each imported contract becomes a draft in OMB Cloud, linked to the client (auto-matched by name in the document), with all extracted data ready for editing. You review, correct any mistakes, and either keep as a record or re-send for re-signature if needed.
Tips
- Bulk-import 5-10 at a time, not 50. Each runs ~30-60 seconds; smaller batches make errors easier to triage.
- For very long contracts (50+ pages), expect the AI to take longer and possibly mark sections as
[illegible]in the densest legal boilerplate. Quick proof read fixes those. - The original file stays attached to the contract record for reference, regardless of what was extracted.