When a page promises a monthly income from an AI side project, you rarely learn where the number came from. Kairo does the opposite and tags every project in its database with a reliability level, so you can tell a figure that rests on sources from one that is only a first sketch.
There are three levels, in descending order of confidence: verified, estimate and hypothesis. A verified entry carries the most confidence, an estimate is a working figure, and a hypothesis is the lowest level: the project looks plausible and the numbers are a sketch.
Of the 14 projects in the database, 5 are verified, 5 are estimates and 4 are hypotheses. That is a fairly even split, and it is the honest one: a database that marked everything as verified would be claiming more than it knows.
What does this mean when you compare two projects? A hypothetical project with an attractive profile, such as WhatsApp Business setup for shops, is worth looking at and worth testing in a small way before you plan around it. A verified entry, like the one for printable books on Amazon KDP, tells you about the downside as well: the data records a median income that is very low for a single title, and a rule requiring you to declare content made with AI.
It also helps to keep the other limits in mind. The figures describe typical cases and not your case. They do not include your local market, your price, the clients you already know or how much time you really have. The tax side is a separate question that depends on where you live and on your situation, and it is worth asking an accountant before you invoice anyone.
The practical use is simple. Use the ranking to narrow 14 options down to three. Then take the reliability label seriously: spend a little time testing the hypothetical ones, and spend more time double-checking the numbers on the verified ones against your own area.
To see which projects rank at the top for your own profile, you can run your profile through Kairo: https://ai-income-matcher.kairo-app.workers.dev
