Didit Console

Manual Review

The manual review interface allows your team to assess edge cases where automated verification may raise warnings based on your configured workflow.

πŸ–₯️ Overview of Verifications

Verifications Overview

In the screenshot above, you can see the verification table view, where each session is listed with its status, document type, country, and more. Verifications can have the following final statuses:

  • βœ… Approved
  • ❌ Declined
  • 🟠 In Review β€” Requires your manual attention
⚠️

Important: Verifications marked as In Review have triggered one or more warning signals and require manual review before a decision can be made. You can find the count of pending reviews at the top of the page.


πŸ” Conducting an Effective Review

Step 1: Initial Assessment

Clicking a session opens a detailed modal view like the one below:

Session Modal
  1. Review warnings (e.g., low liveness score) in the bottom left section.
  2. Check if the user has previous attempts (which may indicate fraud attempts or genuine retries).

Step 2: Document Inspection

  1. Inspect visual security features: holograms, UV patterns, watermarks.
  2. Validate field-level consistency across:
    • Document number
    • Date of birth
    • Expiry and issue dates
    • Name fields
  3. Confirm image clarity and data legibility.
πŸ’‘

Tip: Hover over document images in the Didit console to use the zoom feature for detailed inspection.


Step 3: Biometric (Face) Verification

  1. Carefully compare selfie and document photo:
    • Eye shape and color
    • Nose and mouth geometry
    • Chin and jawline
    • Presence of moles, scars, or tattoos
  2. Account for age progression between document photo and selfie.

Step 4: Risk & Signal Review

Navigate through the AML and IP sections to check contextual risk factors:

AML Screening

AML Tab
  • No hits = Low risk
  • Review all matched entries carefully (PEP, watchlists, etc.)

Location & Device Intelligence

  • Compare document address, IP geolocation, and device metadata
  • Check for use of VPNs, proxies, or emulators

βœ… Making a Decision

Approval Criteria

  • βœ… Document is authentic and readable
  • βœ… Strong biometric match
  • βœ… No relevant AML or fraud signals
  • βœ… Consistent device and location data

Decline Criteria

Decline Modal
  • ❌ Tampered document suspected
  • ❌ Selfie does not match document photo
  • ❌ High-risk AML match
  • ❌ Device/location inconsistency or suspicious behavior
πŸ“

Reminder: Always add a reason and include review notes when declining a session. This supports QA, future audits, and analytics.