# Built with AI: from prototype to launched product > You built an app with AI and it works on your machine. A fixed-price Launch Gate Audit tells you what stands between it and real users. - URL: https://envisageapps.com.au/built-with-ai/ - Type: page ## The wall people hit You built something. It works on your machine. Then you tried to make it a product, and the last stretch started costing more than the first four weeks did. Nothing on this list responds to better prompting. ### The loop Every fix breaks something else. You've spent more on credits going in circles than the build cost, and you can't tell whether you're close or nowhere near. ### The unknown unknowns The code runs. Whether it's safe to put a stranger's email address into it is a different question, and you don't have a way to answer it. ### The first real user It works for you because you use it the way you built it. Nobody else will. ### The gate App Store review, a customer's security questionnaire, an investor's technical review, or the TGA. Each one is a human deciding, and none of them accept "it works". ## What we do about it ### Tell you where you actually are A fixed-scope audit that separates launch blockers from after-launch work: $3,500–$5,500 AUD, five business days. You get a decision, not a document, and a fixed-price quote for the fixes it scoped. ### Fix the classes of bug generated code repeats Auth, session handling, access control. They aren't random; they're the same handful every time, because generated code learned from code that had them. ### Prove it on hardware that isn't your laptop Real devices, old handsets, bad connections, real load. The failures that only appear outside the environment the code was written in. ### Take it through the gate App Store review, security questionnaires, TGA classification. Twenty-five years of reading these letters, and the reason a reviewer gives is rarely the real one. Every finding in the report traces to a file, a test, or something we observed on a real device — never "AI code is bad". A senior engineer validates every finding before it reaches you, and the walkthrough call is where you check us on it. ## What we don't do We won't treat the code as suspect because AI wrote it. Most of what we work on now was built this way. We treat it as inherited. We won't rebuild from scratch unless the audit shows a rebuild is genuinely cheaper — and if it is, we'll tell you in writing, with the numbers. And we won't guarantee approval. Nobody can, and an approval guarantee is a representation nobody can control. What we offer instead is scoped: if the store rejects work we performed, we keep working at no additional cost, or we tell you in writing why it can't pass. ## Where this stops If you're still changing what the product is every week, don't hire us yet — you'd be paying to harden something that's about to change shape. Keep prompting; it's the right tool for that stage. Come back when the shape has stopped moving. If you have already hit a wall and want to know what it was before you talk to anyone, the most common one is written up in full: [why Apple rejects apps under Guideline 4.2](/blog/app-rejected-guideline-4-2/), including the three ways out and the [wrapper tax](/lexicon/#wrapper-tax) you pay by picking the wrong one. ## Common questions ### I built it with Lovable, Bolt, Cursor or Claude Code. Will you work on it? Yes — that's most of our current work. The tool doesn't change the audit; the same classes of finding come up whichever assistant wrote the code. ### It's a web app, not an App Store app. Does the audit still apply? Yes. The store-review items drop out; the security review, access control, infrastructure and — where it applies — TGA classification remain. Scope is set on the walkthrough call before you pay, and the price band is the same. ### Can't I just keep prompting until it's fixed? For features, often yes. For the gates, no — App Store review, a security questionnaire and a TGA classification are humans making judgment calls, and there is no prompt on the other side of the desk. ### What do I actually get? A two-bucket report — launch blockers and after-launch — a walkthrough call, and a fixed-price quote for the remediation it scoped. The audit fee is credited against remediation within 60 days. ## Find out where you actually are START HERE The audit is the front door, and it prices the second step for you: the hardening is quoted from the findings, never from a guess. ### Launch Gate Audit The fixed-scope diagnostic. You get a decision, not a document. **$3,500–$5,500** AUD fixed · 5 business days _Fee credited against remediation within 60 days_ - App Store review — 4.2, 2.5.2, 5.1.1, and Play's 12-tester wall - Security review — auth, session handling, access control, secrets - Australian regulation — TGA classification screen where it applies - Real devices, not simulators - Two-bucket report: launch blockers and after-launch - A senior engineer validates every finding - Walkthrough call, and a fixed-price quote for the remediation [Book a Launch Gate Audit](https://cal.com/nchallen/book) ### Production Hardening The remediation the audit scoped. Quoted from the findings, never from a guess. **from $30,000** AUD fixed _Scoped by the audit before it is quoted_ - Native work where the wrapper fails — push, offline, Keychain and Keystore - StoreKit with server-side receipt validation - Submission managed inside your developer account - 30-day post-launch fix warranty [Book a call](/contact-app-developer/) _If the store rejects work we performed, we keep working at no additional cost — or we tell you in writing why it can't pass. We do not guarantee approval, and an approval guarantee is a representation nobody can control._