Our evidence
Every gap shows what backs it
Anyone can say "you should add a reminder". Pixatom tells you why: which published rule asks for it, which well-known apps already do it, or that it is good practice still worth testing.
A library of real flows, studied screen by screen
Pixatom's evidence comes from a hand-picked library of real product flows: sign-ups, onboarding, free trials, checkouts, upgrades, failed payments, cancellations, sign-in and security, invites, sharing and more. That is more than 150 real flows and over a thousand screens from apps people use every day, each written up step by step: what the app asks, what it shows and what happens next.
When Pixatom says a gap is "seen in" an app, it is pointing at one of those real flows, not at something the AI half-remembers.
Three kinds of evidence
- Published rules. Laws and platform rules that apply to the flow, like subscription rules on reminders and cancelling.
- Real apps. Well-known products that handle the same step well, from the library.
- Good practice. Sensible advice that is not proven for your case. These findings say so, and suggest a test.
How we keep findings honest
- Checked against the library. If an AI step names an app or flow that is not in the library, it is removed before you see it.
- Confidence matches the evidence. A gap backed by one app cannot claim to be certain. Confidence is capped to what the evidence supports.
- A second review. Every gap is checked again by a separate review, which keeps, downgrades or drops it.
- You decide. You agree with each gap or mark it not relevant. Only agreed gaps go into your blueprint.
What we do not claim
Pixatom is a second opinion, not legal advice and not a replacement for talking to your users. When it mentions a rule, check how it applies to you. When it suggests a fix, test it. That is why every blueprint ends with a test plan.