signzy

API Marketplace

downArrow
Logo
Responsive
Decorative line

Document Verification

Overview

Document verificationcaptures ID images, evaluates authenticity, extracts data with OCR or ICR, and binds the document to the applicant via face match and liveness. It orchestrates quality gates, template detection, and jurisdiction-specific rules; handles exceptions like glare or cutoff; and outputs structured attributes for screening and onboarding.
Strong programs include PAD-tested liveness, NFC where available, and database corroboration for higher assurance. Decisioning uses thresholds and reason codes to route approvals or manual review. Privacy controls cover consent, redaction, encryption, and retention. Continuous QA samples outcomes to tune models and thresholds, reducing false rejects and review load.

FAQ

How is verification different from authentication?

Authentication checks genuineness; verification adds capture, data extraction, and ownership binding via biometrics and orchestration. Both are needed for high-assurance onboarding.

What improves success and reduces drop-off?

Real-time capture guidance, template auto-detect, NFC reads when possible, and clear lighting and framing prompts, with graceful escalation.

Do we still need external databases?

Yes. Registry and sanctions checks complement documents, raising assurance and reducing residual risk.

How should exceptions be handled?

Use playbooks with acceptable alternates, set SLAs for recapture and review, and record evidence and rationale to support audits.