Do you need access to production or to customer data?
No. We work on a staging or test build, inside test accounts you create for us. If a flow involves money, we use the payment provider's test mode and never real money.
Will this bury my team in noise?
No. It might bury your team in bugs, though. We found 192 in our own app. Every one is reproduced before you see it, the same fault on two platforms counts once, and each carries a severity so you know where to start. What does not reproduce never reaches your tracker.
Will a long list make my team look bad?
No, and it is not a scorecard. Every product has these. Ours had 192 and it was already live. The register goes to you and nobody else, and a bug you found before a customer did is a credit to whoever ordered the run.
Do you fix what you find?
Your team fixes and we retest, unless you add the fix. With that add-on every bug arrives with a patch on a branch and a test that failed before it, for your engineers to review. We wrote one for every bug in our own register.
We already have automated tests. What does this add?
Tests check what someone thought to ask. This looks for what nobody asked: the stale tab that overwrites new data, the decimal comma that multiplies a payment, the restore that loses payment history. Those three are real entries in the register above.
Is this done by AI or by people?
Both. AI agents do the legwork, which is how one engagement can cover every screen. A person checks every report before it is sent, and that person is who you talk to.
What do you need from us?
Read access to your repository, a build we can run, test accounts, and about two hours of your time. Most of what we find starts with reading the code, so the repository is not optional. Beyond that, share whatever says what the product is meant to do: specs, tickets, designs, a list of known issues. One named engineer who can answer "is this intended?" during the run helps most of all.
What happens to our code and data afterwards?
We sign your NDA before we see anything. Your code and build are used for your engagement only and removed when the last retest is done.
What if you find a security problem?
You hear about it first and privately, outside the main report, the same day we confirm it.
How long before a launch should we book?
As soon as you have a build that runs. The delivery date is fixed on the scoping call, and you will want time to fix what comes out before the launch, not after.
Which platforms do you cover?
Web apps, iPhone apps and Android apps. The two exhibits on this page are web and Android only because that is what those two products ship.
