How we think about the responsibility that comes with assessing people.
Workplace Judgement is built to add credible, structured evidence to a recruitment process — not to replace human judgement. These are the principles that shape every part of the product.
A decision-support tool, not a decision-maker
Workplace Judgement never labels a candidate as Hire, Reject, Recommended or Not Recommended, and it never removes anyone from a process automatically. Every score, behavioural profile and interview prompt exists to inform your hiring team's judgement — the decision itself stays with you.
Employers remain responsible for their decisions
Assessment results are one input among several. We ask every employer to consider them alongside interviews, experience, references and any other evidence relevant to the role — not as a single determining score.
Transparent, non-binary scoring
Nothing is marked simply right or wrong. Every candidate rating is compared against a pre-defined expert effectiveness benchmark on a graded scale, and the method — including how behaviours are weighted — is visible to employers before an assessment is issued.
Consistency, described carefully
Where related items in an assessment probe similar underlying judgement, we surface a consistency indicator — Very consistent, Consistent, or Varied across situations. It is never presented as a lie-detection tool or used to label a candidate as dishonest.
No protected-characteristic inference
The platform does not collect and does not attempt to infer age, sex, race, ethnicity, religion, disability, sexual orientation, gender identity, pregnancy, political belief or health status from a candidate's responses. It does not produce personality labels or psychological diagnoses.
Candidate accessibility, considered from the start
Every assessment includes an accessibility-adjustments step before it begins, and the candidate interface is built to the Web Content Accessibility Guidelines (WCAG) 2.2, Level AA — full keyboard navigation, visible focus states, no colour-only signalling, and no unnecessary timed interactions by default.
Privacy by design
Candidate identity, responses and scoring are held as separate data. Retention periods are configurable, and candidate records can be deleted in line with an employer's own data policy. Candidate response data is never used to train unrelated AI systems.
AI assists with language, never with the decision
Where AI is used to help phrase a narrative summary, the underlying score always comes from the deterministic scoring engine — never from a generative model — and AI is never used to decide whether someone should be hired.