← Resources

Tell the story of public numbers.
Checked twice, before you show them.

Budgets, contracts, recovery funds, your sector's statistics: an AI finds the official figure, confirms it on a second source, redoes sums and rounding and draws the chart in 4:5 and 9:16. Every number goes out with source and period, or does not go out.

Checkingtwo sources per numberFormats4:5 and 9:16ForClaude or ChatGPT

What an infographic costs.

Public prices of an infographic, and the out-of-pocket cost of a piece made with the kit.

Range source: venngage.com (updated 27 Nov 2025), read 8 Oct 2026. Bars not to scale.

One question. Two checks. One form.

The question
A reader's, with an answer in a number: their region, their taxes, a natural comparison.
↓
The data
From the official dataset or report, with unit, period, territory and saved proof.
↓
Check A
Your extraction: no aggregate added to its rows, nominal or real said, perimeter read.
↓
Check B
Another official page showing the same number. A program reads it and redoes sums and rounding.
↓
The form
One big number or a comparison, drawn by code in 4:5 and 9:16, outside the bands the apps cover.

Two ways.

Make them with your AI agent, or I make them on your data.

The kit

Do it yourself

$19

→The ready folder: open it with your agent (Claude or ChatGPT), write "make a piece on ..." and it does the rest.
→The double check: page B, sums, rounding, a report for every number.
→Charts drawn by code, in 4:5 and 9:16 from one file.
→The check of the bands the apps cover, also on Canva images.
→15-page full guide, in English and Italian.

The service

I do it for you

€390 €690

→Four questions chosen with you, on your data or your sector's public data.
→Double check with a report for every number.
→Each piece in 4:5 and 9:16, with caption, source and period.
→Source files to redo them. If you have the kit, its $19 comes off.

Launch price.

Before buying, read the free guide.

9 pages: the problem, what an infographic costs, the method, the checks, two real cases, the rules born from mistakes, the limits. Enough to check by hand the next number you publish.

What you are probably thinking.

"Which data can I use?"

Public data (budgets, contracts, recovery funds, official statistics) and your sector's or company's data, with the same method: extraction, confirmation on a second source, report.

"What if there is no second source?"

The number stays blocked. If an official site blocks scripts, check B is done by hand and what was seen is written down.

"Does it make videos too?"

The kit makes charts and captions. For videos from the same checked numbers, your agent can use the open-source motore-video-ai project.

"Do I need Canva?"

No: the charts are drawn by code. If you prefer Canva, the safe-area check works on images exported from it too.

"Do I need to code?"

No. Your AI agent, Claude or ChatGPT, uses the kit: it installs what is needed, finds the data, runs the checks and shows you the report.

Where it comes from.

It is the method I use to run the social content of an open data project on Italian public spending: at least 8 videos published on Instagram, X and Threads, each double-checked. With the same method I prepared the piece on recovery-fund nurseries in Lazio for a regional party: 54.1% still under construction, against 46.0% in Italy.

Read more: data storytelling examples with public numbers.