Apps built with Lovable, Bolt, Cursor, or Claude ship fast — then run with nobody on call. We adopt them and operate them under a written SLA, with one named person accountable.
Request a Health Check
AI tools are excellent at building software and structurally incapable of being responsible for it. The tool that generated your app will not notice that your database is filling up, will not restore last night's backup, and will not pick up the phone when checkout stops working. Most AI-built apps run with no uptime checks, no error alerting, and no one reading logs.
Backups are usually an assumption rather than a safety net — a backup that was never restored is a rumor. Ask yourself: could your data be recovered today? Do you actually know?
Traditional agencies rarely want this work either; they would rather rebuild your app than adopt someone else's codebase. So AI-built apps end up in production with real users, real payments, and real data — and no operator. That gap is not a tooling problem. It is a responsibility problem, and responsibility is the one thing you cannot generate.
An excerpt from the monthly report every client receives — what ran, what broke, what changed, what it cost.
Monthly operations report — sample excerpt
Restore drills are real recovery tests, documented with timings — not a checkbox.
Three engagements that show what adoption looks like in practice — from first Health Check to quiet months of operation. Read the full stories →
Adopted after recurring Stripe webhook timeouts were quietly dropping renewals. Stabilized the queue, added retry monitoring, and put the founder's mind on other things.
0 missed webhooks since adoption · 99.97% uptime over six months
The Health Check found daily backups that had never been restored and couldn't be. Rebuilt the backup path, then proved it: full recovery rehearsed on schedule, documented with timings.
Restore drill: 14 min to full recovery, tested monthly
A founder commissioned their app from an agency and asked for an acceptance check before paying. The report surfaced exposed API keys and two data-loss paths; the agency fixed all three before handover.
3 critical findings resolved pre-payment · report delivered in 4 days
The Health Check tells both of us what state your app is in. If we then agree to adopt it, Managed Operations is the ongoing contract.
$249 flat · delivered within five business days · read-only
A full review of your production setup: hosting, database, backups, secrets, error handling, dependencies. You receive a written report in plain language, every finding rated by severity, with what it costs to fix — including a tested answer to the question that matters most: could your data be recovered today?
The Health Check also works as an independent acceptance check — verdict before you pay a builder's final invoice — and as a Launch Check: send us the app before go-live and we audit security, compliance readiness, and the operations basics in one pass. The report is yours either way — fix things yourself, hand it to anyone, or ask us to adopt the app.
The guarantee: if the report doesn't surface at least one material risk or defect you didn't already know about, you don't pay.
from $149 per month · written SLA · cancel monthly
We adopt your app and run it: uptime and error monitoring with committed response times, daily backups with scheduled restore drills, security patches applied and verified, and bug fixes within your plan's included hours. Every incident gets a written postmortem; every month you get the operations report shown above.
Everything runs in your accounts — hosting, domain, database, payments. We operate; you own. Leaving Ship & Sleep never requires migrating away from us, because nothing lives with us.
Typical stacks: Next.js · React · Supabase · Firebase · Vercel · Netlify · Stripe · Postgres · serverless functions
The Health Check, step by step → Managed Operations in detail →
Four steps. You can stop after any of them and keep everything produced up to that point.
You tell us what the app does, what it runs on, and what worries you. You hear back within two business days — either a Health Check slot or an honest "we're not the right fit". Free, no obligation.
Read-only access, five business days, written report with severity ratings. Nothing is changed in your app during this phase, and the report is yours to keep.
If we both want to proceed: a fixed-price offer to stabilize the critical findings, plus the operations agreement — SLA, coverage hours, plan. Everything in writing before any money changes hands.
We run the app: monitoring, backups, patches, fixes, monthly report. Your contact is the person who signed the SLA.
And the exit, in writing: if you leave — for any reason — you get a complete handover: credentials returned and rotated, runbook and architecture notes included, thirty days of reasonable questions answered at no charge. Making it easy to leave is part of being worth staying with.
An SLA that costs the provider nothing when broken is decoration. Ours pays: miss a committed response time and you receive a service credit of 10% of that month's fee per incident, up to 50% of the month — automatically, without you having to ask.
| Commitment | Care | Standard | Priority |
|---|---|---|---|
| Coverage window | Mon–Fri, 09–18 CET | Mon–Fri, 09–19 CET | Mon–Sat, 09–21 CET |
| First response, critical incident | Next business day | 8 coverage hours | 4 coverage hours |
| Uptime checks | Every 5 min | Every 1 min | Every 1 min |
| Backups | Daily | Daily | Daily |
| Restore drill (real test, documented) | Quarterly | Quarterly | Monthly |
| Security & dependency patches | Monthly | Monthly | Weekly review |
| Included fix hours per month | — | 4 | 10 |
| Incident postmortem in writing | Yes | Yes | Yes |
| SLA credit on missed response | Yes | Yes | Yes |
Fail-closed by default. When something is ambiguous — a migration that might lose data, a dependency update that might break payments — we stop and ask instead of guessing. The default action in an unclear situation is the one that cannot destroy anything. This principle comes from operating systems where a wrong guess prices itself within minutes.
Trust is mostly built out of refusals. These are ours, stated before you pay rather than discovered after.
Nobody can honestly promise that, including companies far larger than us. We promise defined monitoring, defined response times, and credits when we miss them.
Coverage hours are printed in the SLA. Selling around-the-clock response we cannot reliably staff would be a lie with a delay on it.
We decline apps handling medical records, self-custodied payment data, gambling, or regulated financial services — those deserve specialized, audited operators. We also decline codebases we assess as unrescuable, and say so in the Health Check rather than bill monthly for managed decline.
We operate, stabilize, and fix. Building your roadmap is a different job with different incentives — an operator who profits from rebuilding your app has a conflict of interest we'd rather not own.
Anything beyond routine patching is proposed in writing first. Your app doesn't get "improved" without your approval.
No fake logos, no purchased reviews, no invented client counts. Trust here is a sequence of verifiable steps: read-only first, contract second, access third.
Least privilege, always: read-only access for the Health Check, operator access only after adoption, and only to the systems named in the agreement. Credentials live in a dedicated vault, never in chat logs or plaintext files. Every access-relevant change is logged in your monthly report. When you leave, access is revoked and confirmed in the handover document.
Ship & Sleep is operated from Germany and works under the GDPR. A data processing agreement (Art. 28 GDPR) is part of every operations contract, not an upsell. Your code and data are never used for anything except operating your app — not for training, not for marketing, not for other clients. This website sets no cookies and runs no trackers; the only data we get is what you choose to send us.
One flat-fee Health Check. Read-only, independent, guarantee-backed — and yours to keep, whatever you decide afterwards.
Request a Health Check — $249If your app needs something these plans don't cover, you get a written quote before anything is billed — never an invoice surprise. The Health Check is credited in full against your first month if you adopt within thirty days. Stabilization work from Health Check findings is priced per finding as a fixed quote, agreed in writing — no hourly meter running.
Care
$149 / month
For stable apps that need a watchful eye.
Uptime checks every five minutes, daily backups with a quarterly restore drill, monthly security patches, next-business-day response. Fixes are quoted separately.
Request intake →Standard
$349 / month
For live apps with real users and revenue.
Everything in Care, tightened: uptime checks every minute, an 8-hour committed response, and 4 fix hours included every month.
Request intake →Priority
$649 / month
For apps where downtime is immediately expensive.
Extended coverage including Saturday, a 4-hour committed response, monthly restore drills, weekly patch review, and 10 fix hours included.
Request intake →Prices in USD, exclusive of VAT where applicable. Monthly plans renew monthly and can be cancelled to the end of any month — no annual lock-in, no cancellation fee, full handover included.
One named lead per engagement — not a ticket queue.
Every Ship & Sleep engagement has a named engagement lead: the person who signs your SLA, reads your monitors, writes your postmortems, and picks up when something breaks. You know their name from the first call, and it doesn't change with the weather. Today that person is the founder — by design: founding clients get the person who built the criteria catalog, not employee number twelve.
Ship & Sleep operates from Konstanz, Germany, and works GDPR-native — a data processing agreement is part of every contract, not an upsell. The roster is deliberately capacity-limited: when it's full, the intake says so instead of degrading everyone's service. And because every adopted app gets a current runbook from day one, we stay structurally replaceable — which is exactly why clients stay.
Yes, structurally: everything stays in accounts you own — hosting, domain, database, payment provider. We are added as an operator with the minimum access needed. Leaving Ship & Sleep never requires migrating away from us, because nothing lives with us.
The stacks AI builders actually produce: Next.js / React with Supabase or Firebase, Vercel or Netlify hosting, Stripe payments, Postgres, serverless functions, and the surrounding glue. If your stack is more exotic, say so in the intake — you'll get a straight yes or no, not a "we'll figure it out" that you pay for later.
Your engagement lead — by name, in the SLA. Coverage hours are contractual and priced honestly; nobody is sold a 24/7 promise that can't be staffed. Every app carries a current runbook, so any competent engineer can act on it — that document is yours from day one, not a hostage.
And if a committed response time is missed for any reason, the SLA credit applies automatically. Unavailability costs us, not just you.
Don't — at first. The Health Check is deliberately read-only: you watch how we work before we can change anything. The report shows you our judgment on your real system; the adoption contract, the DPA, and the SLA schedule show you our obligations in writing. Trust here is a sequence of verifiable steps, not a testimonial wall.
You can, and for small things you should. But an AI session doesn't watch your app at night, doesn't rehearse your restore, doesn't carry a response-time obligation, and pays nothing when the fix breaks checkout. We use AI heavily in our own work — the difference you're paying for is that a named person verifies the result and is contractually on the hook for it.
Notice by the end of any month, effective end of that month. You receive the handover package: credentials returned and rotated, runbook, architecture notes, open items, and thirty days of reasonable questions at no charge. No retention offers, no dark patterns.
Tell us what you built and what worries you. You'll hear back within two business days — either a Health Check slot or a straight answer that we're not the right fit.
This form composes an email in your own mail program — nothing is stored on this site, no account needed. Prefer to write directly?
The roster is capacity-limited by contract. Founding-client slots are open now — when they're taken, this line will say so.