Ship the next thing without hiring for it.

Fixing an AI-built MVP usually starts with a paid diagnostic from $350, and most rescue fixes then cost between $500 and $4,000. The diagnostic takes 2–3 business days and gives you a written list of what is broken, what is unsafe and what is fine, in order of launch risk. Fixes usually take 3–10 business days. The price depends on how many structural problems there are, not on how many screens the app has. If the work is bigger than our range, the diagnostic says so.

It may be, but AI-built apps often ship with a few predictable gaps that only show up once real users sign in. The common ones are Supabase row-level security that is missing or too loose, API keys exposed in the browser, auth that only works for the first user, and payment status that can be changed from the client. Our diagnostic checks those and gives you a plain yes, no or not-yet. Using Lovable, Bolt or Cursor is not a problem in itself. It just tells us where to look first.

Usually the app is not receiving or processing Stripe’s webhook events correctly. Common causes are a webhook endpoint set up in test mode but not live mode, a failing signature check, a missing event type such as subscription updates or failed invoices, or a handler that times out and never retries cleanly. We trace the events in your Stripe dashboard, fix the handler, test plan changes, trials, renewals and failed payments in test mode, then confirm in live. Most billing fixes cost $250–$1,200 or $600–$3,500 for larger integrations.

Fix it in almost every case. Most MVPs, including AI-built ones, need targeted repair rather than a rewrite. A rebuild throws away working features and usually reintroduces bugs you already solved. We look at whether the data model, auth and deployment are sound. If they are, we fix the problems in place and leave notes so your team or AI tool can carry on. If a rewrite really would be cheaper, the diagnostic will say so and explain why. Full rewrites into a different stack are outside what we do.

Yes. Supabase auth and row-level security fixes are among the most common jobs we do for startups. We review every table’s policies, test them with more than one user account, lock down sensitive columns such as plan or subscription status, and move any service-role keys out of the browser. We also work with Firebase rules and auth providers such as Clerk. The fixes are committed to your repository with a short note explaining each policy, and most land within a 3–10 business day rescue job.

Usually yes, because a separate marketing site lets your team edit pages without touching the product codebase. Common setups are Webflow, or a lightweight Next.js site with a CMS such as Sanity, with the app on its own subdomain. We keep sign-up, login and analytics working across both, so tools like PostHog still see the full funnel. A landing page starts from $900 and takes 1–2 weeks. A small site of up to five pages typically costs $2,500–$4,000 and takes 2–4 weeks.

Nobody can guarantee that an AI assistant will recommend your product, but you can make your marketing site easy for those tools to read and describe accurately. That means pages rendered as real HTML rather than only in client-side JavaScript, plain descriptions of what the product does and who it is for, a visible pricing page, public docs, and crawlers that are not accidentally blocked. A technical SEO audit starts from $600 and takes 3–5 business days. Positioning and content are yours to decide.

A startup collecting personal information should generally have a privacy policy, collect only what it needs, and send marketing emails only with consent and a working unsubscribe under the Spam Act. Whether the Privacy Act applies to you in full depends on factors like turnover and what you do, so check with your adviser. This is not legal advice. On the technical side we separate transactional and marketing email, record consent at sign-up, keep secrets and personal data out of logs and make onboarding usable by keyboard and screen reader.

No. We do defined jobs rather than embedded, ongoing development, and we do not work for equity. That suits early teams with more defined tasks than developers: a billing bug, an onboarding flow, a pre-launch review or a marketing page. For a steady run of small jobs, a pre-paid block of small jobs usually costs between $500 and $2,000. If you need a full-time engineer, a fractional CTO or a product built from scratch over months, you need a hire or a larger team, and we will say so.

We do not need your passwords or keys, and yes, we will sign a reasonable NDA. Never send secrets through the site. Once a job is agreed, you invite us to your GitHub, Vercel, Supabase or Stripe accounts with the access the job needs, and you remove it afterwards. We test against staging or test mode wherever possible, so most jobs do not need sensitive production data. The code, commits and accounts stay yours, with nothing locked to us.