Headquarters: Melbourne, Australia, with a New York office
Assessed: Skills assessment and job simulation platform with machine-learning scoring of candidate work.
Rubric: version 1. Search protocol applied and all sources accessed: 2026-09-22.
Documented 2, partial 2, absent 4, of eight criteria. This records what is published about the validity and fairness of the scores. It is not an assessment of the product, the model or the underlying science.
| # | Criterion | Level |
|---|---|---|
| C1 | Technical manual or validity report publicly available without a sales conversation | A |
| C2 | Validity coefficients reported with sample size and uncertainty | A |
| C3 | Adverse impact reported by named subgroup | P |
| C4 | Scored construct named and defined, not described only by outcome | P |
| C5 | Evidence of comparability across languages, versions or delivery modes | A |
| C6 | Human oversight documented, with a candidate route to contest a score | D |
| C7 | Independent review named and the reviewer identified | D |
| C8 | Reliability reported, and reported by group | A |
C1. Technical manual or validity report publicly available without a sales conversation
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
C2. Validity coefficients reported with sample size and uncertainty
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
C3. Adverse impact reported by named subgroup
Partial.
A bias audit page states that the audit "assessed whether Vervoe’s assessment scoring produces disparate impact across ethnicity/race and sex categories in accordance with NYC Local Law 144 requirements", with the outcome given as certification.
Source: https://vervoe.com/ai-bias-audit-results/ (accessed 2026-09-22).
The categories analysed are named and the audit is described as repeated at every model release. The page summarises the outcome and publishes no impact ratios or selection rates.
C4. Scored construct named and defined, not described only by outcome
Partial.
"Scores how candidates actually do the work – video pitches, written plans, live role-plays", with skill groups named such as "Discovery Questioning", "Written Communication", "Objection Handling" and "CRM Proficiency".
Source: https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
Skills are named at a job-relevant level of detail. No construct definition is published.
C5. Evidence of comparability across languages, versions or delivery modes
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
C6. Human oversight documented, with a candidate route to contest a score
Documented.
"Recruiters can see exactly what drove each score, override it with human judgment, and trace every shortlisting decision back to the evidence."
Source: https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
Oversight is documented concretely, with explanation, override and an audit trail. The route is the reviewer route rather than a candidate one.
C7. Independent review named and the reviewer identified
Documented.
"Independently validated by Holistic AI. Fairness results published openly, not hidden behind an NDA." The auditor is described as "an independent, accredited AI auditing firm" and the audit as repeated: "Every model release goes through the same audit."
Source: https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
A named external auditor, a stated scope and a stated cadence. No date for a specific audit was found on the pages examined.
C8. Reliability reported, and reported by group
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://vervoe.com/assessments/ai-scoring/ (accessed 2026-09-22).
Absent records the outcome of the published search protocol on 2026-09-22. Where a source is supplied that the search did not reach, the cell is re-scored under the corrections policy, with the date of the change recorded and the previous level retained.
Back to the index, the rubric and the methods note