Nobody hands you a compliant methodology.
So we build yours — and it holds up when someone asks.
Two new statutory duties, not one
- The annual report to the Ministry of Labour
- Answers to employee pay-information requests
Built for every level of user
- Executive summary for the board
- Scorecards and detail for the HR team
Demo, a. s. · Dashboard
Pay transparency — 2026 reporting period
1 Aug – 31 Dec 2026 268 employees Published 31 Aug 2026268
126 F / 142 M3.8%
company level2
of 6 · 64 people€39,780
annualisedFive factors, not the four everyone built for.
§ 3(1) puts mäkké zručnosti — social and communication ability — alongside the four the directive names. A job evaluation built on the EIGE four-factor toolkit does not meet the Slovak text.
Seven figures, and the evidence behind every one of them.
The report the Ministry of Labour expects, produced from your own pay data rather than assembled by hand in the spring.
Every figure traces back to a source system — payroll, HR master, job evaluation. Nothing is keyed in twice, and nothing is a number somebody remembers agreeing.
That traceability is not presentation. It is what lets you answer the question that follows the filing.
The seven figures, computed.
Gender pay gap — mean
Ordinary salary, all employees
Gender pay gap — median
Ordinary salary, all employees
Gap in variable components — median
Bonus, commission, allowances
Share receiving variable components
Men / women
Distribution across pay quartiles
Q1 lowest → Q4 highest, share of women
Every figure traces to source: payroll, HR master, job evaluation. Nothing is keyed in twice. Illustrative — a 268-person employer.
A number you can trace is a number you can defend. That is the difference between a report and a filing.
The filing is annual. The employee's question arrives whenever it arrives.
An employee has the right to ask how their pay compares with colleagues doing comparable work. That right does not end when the report is filed — it stands.
GenEq answers it from the same model: their pay against the median, quartile and range of their peer group, at grade, cluster or category level. The non-disclosure rule at § 6(1)(b) is built in — a peer group that would reveal a specific colleague’s pay is not shown.
Requests are logged, so you know what was asked, when, and what was answered.
One page per employee, off the model that files the report.
Peer group: S4 · 9 people
0.95
22nd
−5.2%
Peers can be read at grade, cluster or category level, because the right comparison depends on what is being asked. Illustrative.
The answer an employee gets and the figure you file come out of one model, so they cannot contradict each other.
There is no implementing decree. The methodology is yours.
Which means it has to be built to be defended, not built to be filed.
§ 14(4) obliges the Ministry to develop analytical tools and methodologies. That is guidance, not binding regulation — nobody is going to hand you a compliant template, and there is no safe harbour in following one.
And the Slovak text is not the directive. § 3(1) puts soft skills — social and communication ability — alongside complexity, responsibility, effort and working conditions. A job evaluation built on the EIGE four-factor toolkit does not meet it. Anyone who bought the off-the-shelf answer already has a gap in their defence.
The dates are the other half of it. Pay structures were due on 31 July 2026, and the first reporting period is running now.
The obligation, dated.
Act in force
Pay ranges in job advertisements, no questions about pay history, gender-neutral job titles
Pay structures due
Aligned to the statutory evaluation criteria: complexity, responsibility, effort, working conditions — and soft skills, which the directive does not name
First reporting period begins
The first report covers 1 August to 31 December 2026
First report to the Ministry of Labour
Seven prescribed figures, for employers with 150 employees or more
Annual cycle begins
250 and above report every year on the full preceding calendar year; 150 to 249 report every three years. A different deadline and a different period from the first report
Verified against the published text of the act, 29 August 2026. Legal interpretation rests with your counsel.
There is no template to fall back on, so the quality of the methodology is the quality of the defence.
A job evaluation, categories you can justify, and a model that produces the figures.
The job evaluation. All five statutory factors, applied consistently across every role, with the reasoning written down.
The categories. Groups of equal work or work of equal value, drawn on that evaluation — and defensible, because the categories are where the reporting either holds or falls apart.
The governed model. Pay, grade, tenure, hours and entity reconciled into one model that produces the seven figures as an output. Not a spreadsheet somebody rebuilds each year.
The documentation. What was decided, on what basis, by whom. This is the part that matters if the methodology is ever questioned, and the part nobody has when the report was assembled by hand.
One model, from payroll to the filing.
Read-only by default. The model sits in the client's own environment, and the evaluation scores are an input to it — which is what makes a category defensible.
The report is due once a year. The questions arrive whenever someone asks — so this is a model you keep.
Pay is usually fair inside a band. The gap comes from who is in which band.
The two figures that are visual say more than the five that are not.
Once you can see the distribution, the mean gap stops being a single frightening number and becomes two ordinary management questions: who gets promoted into the upper quartiles, and who is put forward for variable pay.
Those have answers. They are not payroll errors, and they are not fixed by adjusting salaries at random.
The shape underneath the seven figures.
Quartiles are the employer's own pay distribution, split into four equal bands by headcount. Illustrative.
Progression and reward are management decisions. The pay gap is what those decisions add up to.
The company number can pass while two categories do not.
Reporting is by category of worker, not by company. An employer who looks only at the top-line figure can file a number that passes and still be sitting on categories that breach.
In the illustration, two categories carrying 64 people sit above the threshold while the company figure passes at 3.8%. Management sits at 4.6% — close enough to cross on next year’s hiring alone.
That is not a reporting problem you can catch in May. It is a reason to hold the data at category level all year, where you can see a category drifting toward the threshold before it crosses it.
The headline passes. Two categories do not.
The dashed line is the 5% trigger for a joint pay assessment. Scale runs 0-10%.
Categories are the employer's own, grouped by equal work or work of equal value under paragraph 3(1).
Every category that breaches was drifting for a year before it did. The filing will not tell you that.
The data it takes to file properly is the data that lets you manage pay.
And next year is cheaper than this year.
A defensible methodology only exists once pay, grade, tenure and job evaluation sit in one place. Once they do, they answer questions the report never asks: what a promotion actually costs, where the retention risk is, and what next year’s pay round does to the gap before you approve it.
That same layer is the base of our platform’s HR and people module — pay cost, utilisation, absence, turnover. Turning it on later is configuration, not another build.
Closing the gap is a budget line like any other: costed, spread across pay rounds, and falling as the structure holds. The methodology stays too — the second filing is the output of a model that already exists.
What closing it costs, as a number you can budget.
| Sales12 affected · 9.1% unexplained | €21,490 |
| Customer service19 affected · 6.8% unexplained | €18,290 |
| To clear both breaches | €39,780 |
| Management — optional9 affected · 4.6%, under the trigger today | €14,160 |
| Including management | €53,940 |
Annualised uplift to bring the unexplained portion to zero for affected employees. Phaseable across pay rounds, and it falls once the structure holds. Illustrative.
The obligation recurs every year, and so does the model. The second year costs less than the first.
Inside your own tenant, under your own policies.
The platform goes to the data, not the data to the platform.
Pay data broken down by sex is the most sensitive employee information most organisations hold. For a large or public-sector employer the first question is not what the reports look like — it is whose infrastructure they run on.
Yours. Built natively on Azure and deployed into your own subscription: your tenant, your region, your ownership. There is no shared database where your payroll records sit next to somebody else’s.
If your security review has requirements we have not met, they are requirements, not exceptions.
Nothing is copied to a vendor to be processed.
Azure
Your subscription, your region
Azure SQL
The model and the data stay here
Power BI
Reports and scorecards- Entra ID and conditional access
- Role-based access, row-level security
- Your key management
- Your logging and retention
Read-only by default, and reversible: the exit is a standard SQL export with documented mappings.
What you keep is a governed model in your own environment — not a dependency on us.
Three phases, and the first one tells you whether you have a problem.
Baseline
One pay-period extract. The seven figures computed, the categories that breach identified, and what closing them would cost. Enough to know whether this is a filing exercise or a remediation programme.
Build
The job evaluation across all five factors, the categories drawn and justified, the governed model, and the documentation that stands behind the methodology. This is the work.
Run
The annual cycle, and the data held at category level between filings — so a category drifting toward the threshold is visible while there is still time to act on it.
Start with a baseline on your own data.
One pay-period extract is enough to produce the seven figures, the categories that breach, and what closing them would cost. Then you decide what to do about it.