Read identity documents for KYC onboarding
Know-your-customer onboarding starts with an identity document. PDF Vector reads passports, driver's licenses, national IDs and residence permits from photos or scans, so your team stops typing names, numbers and dates by hand.
Identity Parse turns the document into structured text including MRZ data, Identity Extract returns exactly the fields your KYC process needs, and Identity Ask answers checks such as whether the document has expired. PDF Vector reads documents; liveness checks, sanctions screening and the final verification decision stay with your KYC provider and team.
Where PDF Vector fits in onboarding
- 1
Collect the document
Your onboarding form or app uploads a photo or scan of the ID and sends it to PDF Vector as a file, base64 or URL.
Identity Parse API - 2
Extract the fields you need
Define name, date of birth, nationality, document number and expiry date in a JSON Schema and get consistent JSON back.
Identity Extract API - 3
Run document checks
Ask whether the document is expired, or whether the name matches what the customer typed, and route mismatches to manual review.
Identity Ask API - 4
Keep data in region
Enterprise dedicated instances run in the location you choose, with OCR and AI processing kept inside UK, EU, US, Japan, Australia or Asia Pacific residency zones.
Enterprise instances
APIs for this workflow
Free tools and automation
Frequently asked questions
- Which identity documents are supported?
- Passports, driver's licenses, national ID cards and residence permits, from photos, scans or PDFs.
- Does PDF Vector verify that a document is genuine?
- No. It reads the document and returns its data, including MRZ fields and expiry dates, so you can run checks. Fraud detection, liveness and the verification decision remain with your KYC process.
- Can identity data stay in the EU or UK?
- Yes, on Enterprise dedicated instances, which run in the location you choose. The shared General Instance runs in the US.
- Are ID images kept?
- Files are never stored permanently or used for training, and data is encrypted in transit and at rest. Files from failed requests on the shared General Instance are kept encrypted for up to 14 days to diagnose the failure, then deleted automatically.