Irish employers with 50+ staff publish gender pay gap figures this November

Your gender pay gap report,
done in an afternoon — not a fortnight.

Upload a payroll export. Parity computes every statutory metric, groups roles into categories of equal work, and drafts the written narrative — the part HR actually dreads. You review, approve, and download a submission-ready report pack.

Book a walkthrough

No signup. Sample data. Nothing persists between sessions.

The obligation is already live — and it's about to get bigger

Since the 2025 amendment regulations, every Irish employer with 50 or more employees must pick a June snapshot date and publish gender pay gap figures five months later — a real, recurring November deadline under S.I. 264/2022.

The EU Pay Transparency Directive layers on category-level reporting and a joint pay assessment trigger at a 5% unexplained gap. Ireland's transposition is delayed, not cancelled — employers who wait will be reporting under new rules with old tooling.

What that means in practice

Seven statutory metric families, each with a precise legal definition — mean and median gaps, bonus, BIK, part-time, temporary, quartiles.
A written explanation of why the gap exists and what you're doing about it — published under your name.
Spreadsheets get the arithmetic done. They don't draft the narrative, keep an audit trail, or flag which categories cross the 5% line.

Four steps. Under three minutes in the demo.

1 · Upload

Export payroll as CSV or XLSX and drop it in. Parsed in memory — individual rows are never stored.

2 · Map & categorise

Columns map automatically. AI proposes categories of work of equal value with a written rationale — you confirm every judgement.

3 · Review metrics

Every statutory metric with n-counts and its legal source, quartile distributions, and the 5% flag by category.

4 · Report pack

A drafted narrative specific to your numbers, an approval flow with an audit trail, a branded PDF and a submission-shaped CSV.

Anyone can compute a mean.
Nobody wants to write the explanation.

The written narrative is the part of GPG reporting that actually costs HR sleep: a defensible, published account of why the gap exists and what the employer is doing about it. Parity drafts it from your computed aggregates — specific to your categories and your figures, never generic filler — and hands it to you in an editor. Nothing is published until a person approves it, and every approval is recorded in an audit trail.

Sample draft — flagged category Draft — requires review

In Engineers, the mean hourly gender pay gap is 11.54%. The gap is driven primarily by the distribution of men and women across roles and seniority within this population. The employer is reviewing pay bands, hiring practices and progression criteria for this group, and will reassess at the next snapshot date.

Payroll data is the most sensitive file in the building.
We built for that first.

Raw pay data never reaches the AI

Only aggregates — category names, headcounts, computed percentages — are ever sent to the model. The boundary is a single class in code, and a test fails the build if row-level data could cross it.

Ephemeral by default

Your source file is destroyed the moment processing completes — the deletion timestamp is shown on screen. We keep the computed aggregates and the report. Nothing else, unless you opt in.

EU hosting only

Hosted on EU infrastructure (Frankfurt, Germany). No US data transfer for your payroll processing. Human confirmation on every AI judgement, by design.

Metric definitions live in configuration with per-metric legal citations — when the Irish Pay Transparency Bill lands, Parity updates by definition change, not by rebuild.

HR consultancy or accountancy practice?

If you run GPG reporting for a portfolio of 50–250 employee clients in spreadsheets today, Parity is available for licensing and white-label. One conversation could take this November's reporting season off your bench.

Talk to us about partnership

Questions we expect you to ask

Does our payroll data train an AI model?: No. Individual salary rows never leave the application at all — only computed aggregates are sent to the model for categorisation and narrative drafting, over the Claude API. Your data is not used for model training.
What do you store?: By default: the computed aggregates, your confirmed categories, the narrative and the report. The uploaded file itself is deleted the moment processing completes — you can see the deletion timestamp in the app. Retention of source files is strictly opt-in.
Do you submit to the portal for us?: No — you get a submission-shaped CSV and a formatted PDF, and you file them. The national portal's rollout is still moving; producing a clean file you control is the dependable path this cycle.
What does it cost?: Pricing is per annual reporting cycle and depends on headcount and whether you're an employer or an adviser running multiple clients. Email d@devensis.com and we'll quote you the same day.
What happens when the Irish Pay Transparency Bill lands?: Parity's metric definitions are configuration with legal citations, not hard-wired code. The Directive-level view ships today as a clearly separated preview; when the Bill finalises the rules, the definitions update without a rebuild.

Your gender pay gap report, done in an afternoon — not a fortnight.