▚
cpuPulse cpupulse.app
See what every PC actually runs, hour by hour. A tiny Windows service samples CPU per app and draws a daily heatmap — never screens, never keystrokes, never URLs.
$1 per PC / month
Software studio · Portugal · since 1996
Small software, built to be finished.
We build one small product at a time, do one job properly, and charge a dollar a month for it. When a product is finished, the next one starts.
Thirty years of ID-card software before this. Same company, shorter loop.
Products
Two priced. Three being built.
A price appears here only when the product is decided. Nothing else carries a number yet, because a number we would have to change is worse than no number at all.
◍
Privalytics privalytics.app
See your visitors on a map, not in a database. A cookieless script and a calm daily picture. Counts only — no row per visit, ever. We count, we don't track.
$1 per site / month
In development · pricing not decided
Sentrel
Know before it breaks. Uptime, and the part most tools forget — certificates, domains and tokens that expire on a date nobody is watching.
Granvard
Security posture for small companies, with nothing installed on any machine. Findings in plain language, and the report an insurer or auditor actually asks for.
Metis
A company assistant that answers from your own tools, with proper isolation between tenants and an audit trail on every call.
Smallest first, largest last. A product only starts when the one before it is done.
How we build
Written down before it is written.
Every product is a plain-text document before it is code: what it does, what it will never do, what it stores, what it costs. The document is the source of truth, and the build follows it. That discipline is why a small studio can ship finished software — and why a dollar a month is enough to sustain it.
01
Decide
One job, one price, and a written list of what the product will never do. Settled before anything is built.
02
Specify
The data, the rules and the limits, in plain text. No slide deck, no ticket, no diagram to go stale.
03
Build
Modern tooling — AI included — does the typing. Every line is reviewed, tested and signed by a person before release.
04
Finish
Ship it, support it, stop adding to it. A finished product is what frees us to start the next one.
Four steps, then again — the mark is the process.
Work with us
You have the idea. We build it.
The same loop works for someone else's product. A small number of projects a year, taken one at a time, specified before written and finished before the next one starts.
How it works
- →We write the spec together, in plain language, before any code exists.
- →Fixed scope and a fixed price, agreed on that document.
- →You own the code, the domains and the accounts. From day one.
- →It ships finished, with the documentation that explains it.
What we are not
- —Not an agency. There is no team here that needs keeping busy.
- —No retainers, no monthly minimum, no hours to fill.
- —No lock-in. Nothing you cannot take somewhere else.
- —No project we cannot finish. If it is too big, we say so.
Principles
The rules that survive contact with a deadline.
01
One price, worldwide. No tiers, no seats, no feature gates, no "contact sales".
02
Record less than you fear. A product that promises restraint cannot ship one feature that breaks it.
03
Nothing hidden from the person at the keyboard. Named services, visible processes, normal uninstall.
04
Prepaid, never a surprise bill. Credit runs out, the product pauses. That is the whole billing model.
05
Small and done beats big and pending. Each product's scope is closed on day one.
06
A person signs every release. Reviewed, tested and code-signed by the company. That part is not delegated.
About
Thirty years of shipping, now with a shorter loop.
Karditor started writing card-printing software in 1996 and spent two decades building for the global ID-card reseller market — desktop, server and web, for every brand of hardware, never selling to the end user.
That market taught the discipline this studio runs on: small tools, long lifetimes, no lock-in. The tools got faster; the standard did not move. So the products got smaller, the prices got smaller, and there is finally time to build all of them.