Skip to content

THE METHOD

How Krudo assesses - and what it refuses to.

Six independent areas, weighed together. We describe what is assessed, never how - enough to trust the result, not enough to help anyone evade it.

Six areas, judged one by one.

  • Identity consistencyWhether every source describes the same person95
  • Document integrityWhether the CV is what it claims to be88
  • Contact authenticityWhether the details belong to an established person71
  • Device and networkWhat the submission itself reveals62
  • Digital footprintWhether this person existed before today34
  • Cross-applicationWhether this application is acting alone18

THE AGREEMENT RULE

A score only climbs when areas agree.

One ambiguous observation is deliberately prevented from deciding an outcome. A brand-new email domain alone is nothing; the same domain plus a borrowed photo plus no history is a pattern.

STAYS BELOW REVIEW

18/100

  • Document integrity+18
0REVIEW ≥ 40100

One signal alone

RAISES A QUESTION

54/100

  • Identity consistency+22
  • Document integrity+18
0REVIEW ≥ 40100

Two areas agree

DEMANDS ACTION

88/100

  • Identity consistency+22
  • Document integrity+18
  • Contact authenticity+14
0REVIEW ≥ 40100

Three or more agree

THRESHOLDS

Where the line sits.

Nothing is auto-rejected, and nothing is auto-advanced. Below the review threshold, candidates are shown as clear; above it, they wait for a human. The line is the same for every workspace, and it is published rather than hidden.

  • Clear
  • Unresolved
  • Elevated
  • High risk
0REVIEW ≥ 40100

The review threshold is 40, and the four bands turn at 25, 40 and 75. They are the same for every workspace, they are shown here rather than hidden, and Krudo never moves them silently.

Built to be defensible, not just accurate.

Nationality is never scored

A candidate's declared country or name has no bearing on their assessment. Where an application comes from matters only when it contradicts what it says.

Evidence in their favor is weighed

Findings that argue for the candidate are shown on the same scale as findings against them - never quietly discarded.

Same evidence, same result

No prediction, no drift. An assessment made today can be revisited months later exactly as it stood.

Methodology stays internal

Reports describe what was assessed, never how. Nothing published helps anyone evade the checks.

See the method applied to a real report.