Overview

A question frequently raised by lenders and compliance teams is how Begini ensures that the person completing the psychometric assessment is the same person who applied for the loan. This article describes Begini's approach to identity binding, the layered controls available, and how Begini sits alongside, rather than replacing, the lender's existing KYC infrastructure.

Important Boundary: Begini is a behavioural signal layer, not a regulated KYC provider. Identity document verification and regulatory KYC compliance remain the lender's responsibility. Begini strengthens confidence around the assessment session, it does not replace the KYC workflow.

The Five Layers of Identity Assurance

Layer 1 — Application Record Binding

Every assessment session is tied to a specific applicant journey via the lender's own identifiers, application ID, customer reference, phone number, or unique ID. This ensures the psychometric result cannot float unassigned and is always traceable to a specific loan application record.

The lender passes applicant identifiers in the Begini API integration at session initialisation. See Creating a Session for the full parameter reference.

Layer 2 — Verified Channel Delivery

The assessment link is delivered through a channel already associated with the verified applicant, typically their registered mobile number (via SMS) or authenticated in-app session. This creates a possession control: receiving the assessment link requires access to the applicant's verified contact point.

• OTP confirmation before assessment launch adds a step-up layer at point of entry
• In-app delivery within the lender's authenticated session provides the strongest channel assurance
• Multi-device handoffs (e.g. application started on web, assessment on mobile) are permitted but logged and can be configured to lower session confidence

Layer 3 — Liveness Check at Assessment Entry

For higher-risk use cases, a liveness check at the point the assessment is launched provides the clearest identity assurance: a live person is present, and their face matches the identity document on file.

• Begini supports liveness integration at assessment entry via third-party SDK
• The liveness check should match against the ID document captured during the lender's existing KYC process
• No match = no assessment access; assessment is suspended pending manual review

Operational Note: Whether liveness at assessment entry is implemented by the lender's existing KYC platform or a third-party provider is a configuration choice. In most deployments, the lender's onboarding flow completes document and liveness checks before handing off to Begini for the assessment.

Layer 4 — Device and Session Continuity

Where the lender's onboarding journey and Begini assessment occur in the same device and app session, there is a stronger chain between the person who completed KYC and the person completing the game. Begini captures device signals that support this continuity check.

• Device fingerprint at session start is logged and can be matched against onboarding device context provided by the lender
• Significant device or environment changes mid-session are flagged
• For returning borrowers, device consistency between applications is a positive signal; inconsistency can trigger review

Layer 5 — Behavioural Consistency Over Time

For repeat borrowers, Begini can compare interaction style and trait profiles across applications. Large deviations that exceed plausible natural variation are flagged as anomalies — a useful signal for detecting proxy completion or identity reuse across applications.

Responsibility Matrix

Step-Up Verification

When session confidence or identity assurance falls below the lender's configured threshold, Begini supports routing to step-up verification rather than accepting or rejecting the session automatically. Available step-up options depend on the lender's own infrastructure but commonly include:

• OTP reconfirmation to registered number
• Selfie-to-ID face match via integrated liveness provider
• Agent callback or video KYC
• Manual review queue with full session event log

Next steps

• AI-Resistant Assessment Design — how Begini makes AI-assisted completion difficult and detectable
• Fraud Detection & Trust Score — how Begini evaluates session reliability and flags suspicious behaviour
• Creating a Session — pass applicant identifiers at session initialisation