← All use cases
Use case 01

Pay gap and pay equity analyser

Load a workforce file and see two numbers side by side: the gender pay gap across your whole workforce, and the like-for-like gap between women and men doing similar work. Then see why they differ, by WGEA category and by department, and as a sample WGEA-style report.

Vibe-coded tool No AI needed to run it Works offline
Try itUse the real tool. Watch itFive steps, animated. Build itPrompts you can paste. Check the dataWhat is safe, and where. Take itDownload the finished file.

What it does and why it matters

The two numbers answer different questions. The headline gap compares the average pay of all women with all men. The like-for-like gap compares women and men doing similar work. Executives and boards often mix them up.

A large headline gap with a small like-for-like gap usually points to who holds the senior and higher-paying roles, not to unequal pay for the same job. This tool shows both numbers and splits the difference between them.

The like-for-like view can also allow for performance rating, time in role and location, so a gap is not explained away, or missed, because of them. The headline view breaks the gap down by WGEA occupational category and by department, and a sample report lays the results out the way a WGEA benchmark report does, with an invented comparison group. A talking points page turns the numbers into a short brief for executives that you can download and edit.

It was "vibe coded": you describe what you want to an AI assistant in plain English and let it write the code. Once built, it is ordinary code. It runs the same calculation every time and uses no AI when you run it.

1Try it

This is the real tool, running in your browser on an invented workforce. Switch tabs, change the pay measure, or load your own file.

Your file is read by your own browser. Nothing is uploaded.

2Watch it

Five steps from a raw file to a pay gap story. Press Play, or step through at your own pace.

Step 1 of 5Invented data
workforce.csv
employee_idgenderleveljob_familybase_salarytotal_remuneration
E1042WLevel 3Finance$98,500$112,300
E1043MLevel 3Finance$104,200$118,900
E1107WLevel 4Technology$131,000$149,500
E1118MLevel 2People$84,600$93,800
E1190MLevel 5Technology$168,000$201,400

Start with a file. One row per person. No names needed, an ID is enough.

Average base salary, whole workforce
Men
$124,423
Women
$107,860
13.3%headline pay gap

The headline gap. The average pay of all men compared with all women: (men − women) ÷ men. This is the style of figure WGEA publishes for employers.

Women and men compared within the same group
Level 3 · Finance
MenWomen
Gap 1.9%
Level 4 · Technology
MenWomen
Gap 3.1%
Level 2 · People
MenWomen
Gap 0.8%
Like-for-like gap2.6%

Same job, same level. Now compare only people doing similar work. The group gaps are averaged, with bigger groups counting for more.

Splitting the 13.3 point headline gap
Where people sit: 11.7 points Similar work: 1.6 points
Share of women by level
L164%
L258%
L349%
L440%
L531%
L622%
WomenMen

Why the numbers differ. Most of this invented gap comes from who holds the senior roles, not from pay for the same job.

Results by group
GroupWomenMenGap
Level 3 · Finance14121.9%
Level 6 · Legal297.0%
Hidden: small group
Level 2 · People1870.8%

Protect small groups. Groups with only a few people are hidden, so no one's pay can be worked out.

All figures in this walkthrough are invented, for illustration only.

3Build it

You can build your own version with any capable AI assistant your employer allows, for example Microsoft Copilot, Claude or ChatGPT. You describe what you want, it writes the code, and you check the answers.

Build it with invented data, never your real file. Expect two or three rounds of "that's not quite right, change X".

Step 1: Describe the tool

Say what the tool is, what goes in, and who will use it.

I work in HR in Australia and I am not a developer. Please build me a pay gap analysis tool as a single HTML file that I can save to my computer and open in a browser by double-clicking it. Requirements: - It must run offline. Use no external libraries, fonts, images or links, and send no data anywhere. - The user loads a CSV file with these columns: employee_id, gender, level, job_family, base_salary, total_remuneration. These columns are optional, and each one switches on an extra view: department, wgea_category, performance_rating, time_in_role_years, location. - wgea_category holds one of: CEO, Key management personnel, Other executives / General managers, Senior managers, Other managers, Professionals, Technicians and trades workers, Community and personal service workers, Clerical and administrative workers, Sales workers, Machinery operators and drivers, Labourers. - Gender is recorded as W (women) or M (men). Ignore any other value in the calculations, but tell the user how many rows were ignored. - All pay amounts are annualised full-time-equivalent dollars, so do not adjust for hours or part-year work. - Build in an invented sample workforce of about 600 people, with every column filled in, so the tool shows results as soon as it opens. Add a button to download that sample as a CSV so I can see the expected file layout. - The audience is HR people, not data people. Use plain English and Australian spelling (for example "organisation" and "analyse"). Add one short sentence under each result explaining what it means. - Keep the design clean, with large readable numbers. Before you write any code, tell me in a few sentences what you plan to build, and ask me about anything that is unclear.

Step 2: Spell out the calculations

Do not let the AI guess the maths. Tell it exactly what to calculate.

Now add the calculations. Follow these definitions exactly. Do not substitute your own. 1. Gender pay gap (%) = (men's figure − women's figure) ÷ men's figure × 100. A positive number means men are paid more. 2. Show both the average (mean) and the median, for both base_salary and total_remuneration. That is four gap figures in total, shown side by side. 3. Like-for-like gap: compare women and men within the same level and job_family. Calculate the gap for each group using the formula above, then combine the groups into one figure using a headcount-weighted average, so a group with more people counts for more. Show it for the same four measures. 4. Breakdown: show how much of the headline gap comes from where people sit. To do this, re-price everyone at the average pay of their own level and job_family group (regardless of gender), then recalculate the gap using those group averages. That figure is the part explained by where people sit. The remainder (headline gap minus that figure) is the part explained by pay differences within similar work. Show both parts in percentage points, and make sure they add up to the headline gap. 5. Let the user switch between base salary and total remuneration, and between average and median. 6. Adjusted gap: run an ordinary least squares regression of the natural log of pay on a woman/man indicator, level and job_family, plus performance_rating, time_in_role_years and location when the user switches each one on. Report the adjusted gap as (1 − e^coefficient on the woman indicator) × 100, with a 95% range. Show "not enough data" if there are fewer than 40 people, or fewer than 10 women or 10 men. Show how many points of the headline gap each factor explains, and make sure the parts add up to the headline gap. 7. WGEA categories: show the average and median gap for each wgea_category, with an "All managers" line (CEO, Key management personnel, Other executives / General managers, Senior managers, Other managers) and an "All non-managers" line (everyone else). Warn me if the file has more than one CEO. 8. Departments: show one line per department with the average gap, median gap and adjusted gap, and add a department filter that re-runs everything for one department. 9. Sample report: add a page laid out like a WGEA Industry Benchmark Report, covering the headline gaps, the gap by WGEA category, the share of women and men in each pay quartile, and the share of women and men among managers and non-managers. Use invented comparison-group figures that I can overtype, label them clearly as invented, and mark the page "Sample only, not an official WGEA report". Explain each formula in one plain-English sentence on the page.

Step 3: Build in privacy

Pay data is sensitive. Put the protections into the tool itself.

Now add privacy protections. - Hide the results for any group with fewer than 5 women or fewer than 5 men. Show "Hidden: small group" instead of the numbers. Make the threshold of 5 a setting the user can change. - Where you reasonably can, make sure a hidden group cannot be worked out from other numbers on the screen. - In the WGEA category and department views, show gaps only, never average pay, so one person's pay (the CEO's, for example) cannot be worked out by subtraction. When one department is being viewed, leave dollar figures out everywhere. - Do not ask for or display names. Use employee_id only, and do not show individual rows in the results. - Do not save anything. Do not use cookies, local storage or any network request. Nothing may leave the computer. - Show a clear line near the top: "Your file is read by your own browser. Nothing is uploaded or saved." When you finish, list every place in the code that could store or send data, and confirm there are none.

Step 4: Make it prove itself

Ask the AI to give you a test you can check by hand, then check it.

Please help me check that the calculations are right. 1. Create a tiny test CSV of exactly 8 people (4 men and 4 women) with round-number salaries, so I can check every answer by hand. 2. Add a "Show your working" panel to the tool that lists, for the test file, each step of the average gap and the median gap: the men's figure, the women's figure, the subtraction and the division. 3. Tell me what answers I should expect before I load the file, so I can compare them with what the tool shows. 4. If any result does not match a hand calculation, say so plainly and fix the calculation. Do not adjust the test data to make it fit.

A check you can do by hand

Invented example
PersonMenWomen
1$100,000$90,000
2$100,000$100,000
3$120,000$100,000
4$160,000$110,000
Average$120,000$100,000
Median$110,000$100,000
Average gap: (120,000 − 100,000) ÷ 120,00016.7%
Median gap: (110,000 − 100,000) ÷ 110,0009.1%

If the tool gives a different answer, it is wrong. Tell the AI what you expected and ask it to fix the calculation.

Good habits

  • Test with a file where you already know the answer.
  • Ask the AI to explain any formula in plain English.
  • Keep the prompt you used so you can rebuild or audit the tool later.
  • Get someone else to check the numbers before anything goes to an executive or board.

4Check the data

What this touches

Individual pay and gender, which people rightly treat as highly sensitive. An employee ID is enough. Leave names, dates of birth and addresses out of the file.

What is safe

The finished tool runs in your browser and nothing is uploaded. When you build your own, give the AI assistant invented data only. Do not paste or upload real pay data into a public AI tool. Use tools your employer has approved.

Ask before you start

  • Is this AI tool approved for our data?
  • Where is the data stored and who can see it?
  • Does our privacy officer need to know?
  • Who else will see the output?

Privacy rules for employee records differ between private employers and government agencies, so check with your privacy officer. Start with the OAIC's guidance on the Australian Privacy Principles and on employment and privacy. General information, not legal advice.

How this lines up with WGEA

  • WGEA expresses the gap as the difference between men's and women's pay as a percentage of men's pay. This tool uses the same formula.
  • WGEA calculates both average and median gaps, for base salary and for total remuneration. This tool shows all four.
  • WGEA works from annualised, full-time-equivalent amounts, so part-time and part-year employees are compared fairly. Convert your pay figures before you load them.
  • WGEA's comparison is between women and men. Employees recorded as non-binary are not part of that calculation, and this tool treats them the same way.
  • WGEA's inclusion rules have changed over time, for example on CEO pay. Check the current reporting guide for the year you are reporting.

This tool is for exploring your own data. It does not produce or replace your WGEA submission. Its sample report copies the layout of a WGEA benchmark report, but the comparison group and last-year figures in it are invented.

Source: WGEA reporting guide, gender pay gaps

5Take it

Download the tool Open it in a new tab

It is a single file. Save it anywhere and double-click to open it. The sample file can be downloaded from inside the tool.

What next: Back to all use cases