Prototype · ASI Redesign · 2026

A salary review that thinks out loud .

Upload a roster. Watch compa ratios, cohort gaps and EUPTD risks surface themselves — with defensible reasoning attached to every number.

Six acts. One story per employee.

Every salary decision gets a paper trail — not a spreadsheet artefact, but a human-readable rationale a lawyer, a manager and a regulator can all read.

1 Act I · Overture

Upload the roster

Drop a spreadsheet of an entity — e.g. every employee in France. Parsing happens locally in the browser; the file never leaves the device.

2 Act II · The reveal

Compa ratios, instantly

Each person is placed on a map — compa ratio on one axis, performance on the other. Outliers stop hiding in pivot tables.

3 Act III · The cohort lens

Judged against peers, not averages

Click anyone. The system rebuilds their cohort — same job family, grade, country — and shows where they sit against P25, Median and P75.

4 Act IV · EUPTD check

The 5% line, held honestly

When an unexplained gap exceeds the EU Pay Transparency Directive's materiality threshold, it surfaces — with the reasoning that would satisfy an auditor.

5 Act V · The suggestion

Three defensible bounds

Not one number. A Minimum Defensible, a Neutral Median and a Market-Competitive stretch — each one accompanied by the "why."

6 Act VI · Synthesis

Budget reality, staged if needed

Roll-ups by country. Remediation plans staged across two or three cycles when the budget can't absorb the full close. Export ready.

Make the reasoning visible.

Three beliefs shaped every choice in this prototype.

Defensibility over decoration

Every number produces a sentence a regulator could read. The reasoning is the product — numbers are just the artefact.

Cohorts, not averages

Fairness is a peer concept. Every judgement is relative to the right cohort — job family × grade × country — not a global mean.

Motion with meaning

Animation is reserved for moments of understanding — a band drawing itself, a gap filling in. Never decoration.

How we decide the "suggested range."

When the cohort gap crosses 5%, the system proposes three bounds and tells you why each one lands where it does.

Minimum Defensible

Closes the gap to within 5% of the cohort median — the lowest legally-defensible remediation under EUPTD materiality.

Neutral Median

Brings the person to cohort median. Gap fully neutralised; no objective justification required.

Market-Competitive

Pushes toward cohort P75. Reserved for high performers and scarce-skill holders, with justification surfaced.

If the total remediation blows the merit budget, the system proposes a staged plan across two or three cycles — and labels it as a Remediation Plan, not a merit increase. This distinction matters: regulators will ask whether you knew about the gap and what your plan was.