nextpixel

AI readiness assessment

Twelve questions about your data, processes and governance. You get a score, an honest read on where you stand, and the specific things worth fixing first. Nothing is stored and no email is required.

Runs entirely in your browser — nothing is sent to us

01How clearly can you describe the specific problem you want AI to solve?
02Could you tell whether the AI got it right?
03Where does the data this would need actually live?
04What condition is that data in?
05Do you have historical examples of the task being done correctly?
06How consistently is this task performed today?
07What volume are we talking about?
08Do the systems involved have APIs?
09Who can grant access to those systems?
10Who owns the outcome of this project internally?
11What are the consequences if the AI gets something wrong?
12What has to be true before this could touch production data?

0 of 12 answered

How this is scored

Twelve questions across five categories, three points each, normalised to 100. Data and governance carry the most questions because they are where projects most often fail — not because they are the most interesting.

The scoring is deliberately unforgiving at the low end. Telling someone they are ready when their data sits in a system with no API is how AI projects reach month nine with a demo and nothing else.

It is a heuristic, not a diagnosis. A low score does not mean AI cannot work for you — it usually means something unglamorous needs fixing first, and that fixing it is cheaper than discovering it later.

Want the full version?

The paid readiness audit does this properly — interviews with the people doing the work, a hands-on look at your data and systems, and a written report with costed options that you keep either way.

Other tools