Skip to main content

What is a Verification?

A Verification is the result of running checks against an identity document. When you create a verification, Beltic processes the associated document, performs a series of authenticity and data-quality checks, and returns a detailed result. Verifications are the core of the Identity API’s value — they tell you whether a document is genuine, whether the data can be trusted, and whether any issues were detected.

Verification Types

How Verification Works

When you create a verification, the following happens synchronously:
  1. Document processing — The document images are analyzed using advanced document recognition
  2. Data extraction — Personal information is extracted from the document (name, DOB, document number, etc.)
  3. Check execution — Multiple verification checks are run against the document
  4. Status determination — An overall status is computed from individual check results
  5. Response — The complete verification result is returned
Verification creation is a synchronous operation. The response includes the full result — no polling required.

Creating a Verification

With a Session and Document

The most common pattern — verify a document within a session:

Standalone (Without a Session)

You can also run a verification without a session — useful for API-only integrations:

Verification Response

A verification response contains the status, extracted data, and check results:

Verification Statuses

Understanding Checks

Each verification includes an array of checks — individual tests performed against the document. Every check has:

Check Categories

High-level status indicators from the document processing engine:
  • status_optical — Overall optical analysis result
  • status_portrait — Portrait/photo analysis result
  • status_rfid — RFID chip reading result (if applicable)
Detailed analysis of the document’s visual elements:
  • optical_doc_type — Document type recognition
  • optical_expiry — Document expiration check
  • optical_image_qa — Image quality assessment
  • optical_mrz — Machine Readable Zone validation
  • optical_security — Security feature analysis
  • optical_text — Text consistency and validity
Advanced fraud detection checks:
  • authenticity_uv_luminescence — UV luminescence patterns
  • authenticity_ir_b900 — Infrared B900 analysis
  • authenticity_image_pattern — Image pattern analysis
  • And more — the exact checks vary by document type and capture method
Ensure the document images meet minimum quality standards:
  • image_quality_glare — Glare detection
  • image_quality_focus — Focus/sharpness assessment
  • image_quality_resolution — Resolution adequacy
  • And others for colorness, perspective, bounds, etc.

How Status is Determined

The overall verification status is computed from the individual checks:
  • If any check with requirement: "required" has status: "failed" → verification is failed
  • If any check has requirement: "requires_retry" → verification is requires_retry
  • If all required checks pass → verification is passed

Auto-Update Session

When you include "auto_update_session": true in the verification meta, the session is automatically updated with the extracted data from the document:
  • Person name (first, middle, last)
  • Date of birth
  • Sex
  • Address (if present on the document)
This saves you from manually copying extracted data to the session after each verification.

Listing Verifications

Retrieve verifications with optional filters:

Next Steps

For a complete end-to-end example of running a document verification through the API, see Running a Document Verification.