Prototype navigation only — switching does not authenticate across portals or change tenancy. Employer work uses employer sign-in; Operations staff use the Ops staff switcher.
Technical prototype mathematical model — implementation in src/lib/nexus.ts
N = equally-weighted average of all pillar scores P_p
P_p = Σ(R_e,p × C_e,p) / Σ(C_e,p) where R is employer rating (1–5) and C is confidence (0–1).
C_e,p = 0.25·EQ + 0.20·EA + 0.20·VI + 0.25·EDC + 0.10·CRC
EQ = evidence quality, EA = evidence authenticity, VI = verification integrity, EDC = employer data consistency, CRC = cross-record consistency.
Employer Data Consistency = equal average of 5 behavioural factors
Historical accuracy, behavioural consistency, submission quality, update integrity, multi-employer alignment — each 0–1.
Variance of per-employer mean ratings → high / mixed / low
Spread < 0.4 = high, < 0.9 = mixed, otherwise low. Transparency signal only — not blended into hiring verdicts.
Verified evidence ratio → High / Medium / Low
≥75% verified with ≥3 items = High; ≥40% = Medium; otherwise Low. Does not trigger Amber for initial records.
Based on available verified employer assessments, supporting-evidence quality, contributor credibility, employer submission consistency, and consistency across available records. ERN does not make hiring decisions.
Legal concerns, disputes, and serious-conduct flags are never blended into the professional index. They feed only the traffic-light record-status indicator (L7) and the Serious Conduct panel.
To promote fair and relevant hiring decisions, employers generally view employment records within ERN's standard visibility period. Older records may no longer appear in employer searches even though they continue to be securely retained by ERN where lawful, allowing future legal, regulatory, correction, appeal, or policy requirements to be fulfilled.
Visibility periods may change over time in accordance with applicable law, regulatory guidance, and ERN governance policies.