One identity discrepancy can stop an unreliable assessment moving forward.

Build an exception process that distinguishes a flag from a confirmed proxy attempt.

Scenario context

IT services provider · distributed technical hiring · 3,000-interview programme

The problem

A remote assessment feeds a client-facing deployment pipeline. An identity discrepancy needs escalation, but an unusual interaction alone cannot justify calling someone a proxy candidate. The hiring team needs a clear verification and disposition process.

The proposed approach

In a fictional 3,000-interview programme, assume 60 identity-related discrepancies are escalated.
Authorised reviewers clear 42 after reviewing permitted evidence and candidate clarification.
Of the remaining 18, assume six are corroborated proxy attempts and 12 remain unresolved pending further verification.

Where people stay involved

Human reviewers establish identity findings and the customer determines progression. No autonomous facial-recognition feature or automatic rejection is implied.

What the model calculates

The accounting is explicit: 60 escalations equal 42 cleared, six corroborated and 12 unresolved. The modelled escalation rate is 2%; it is not a detection-accuracy statistic.

What it does not establish

These incident counts are invented scenario inputs. They do not support a real fraud-prevention claim, sensitivity/specificity estimate or number of bad hires prevented.

Book a demo · HirePlusPlus home