Guide

How to take an app built with Lovable or Cursor to production.

The tools got you a working product in days. This is the work that is left before strangers can sign up and pay, in the order to do it.

Updated 18 November 20266 min read

The short answer

Keep what works. Rebuild the flow people pay for.

A product built with AI tools is usually worth keeping. What it needs before real users is narrower than a rewrite: a rule for who can see whose data, an answer for what happens when something fails, and one flow that feels finished. Our Product Rebuild does that for one key flow, from $20,000, in 4 to 6 weeks.

Our prices for this work. Each is fixed in a written scope before the work starts.

To find what is wrong firstWorkflow Sprint$5,0005 daysA diagnosis of one problem and a clickable prototype you can test with users.What’s included
To take one flow to productionProduct Rebuildfrom $20,0004–6 weeksOne key flow redesigned end to end, with its components, states and edge cases, built full stack in your codebase.What’s included
To keep going every monthStudio Partnerfrom $5,000a month, design, or from $10,000 a month, design and engineeringOur designers, or our designers and engineers, on an agreed area of your product, with a weekly review.What’s included

What the tools leave to you.

AI tools are good at making a screen that works when everything goes right. Production is everything else. Check these in order.

  • Who can see what

    Every table and every endpoint needs a rule for who may read it and who may change it. Check this first, because a missing rule does not show up as a bug.

  • It is on the public record

    CVE-2025-48757 describes sites generated by Lovable through April 2025 where an insufficient row-level security policy let unauthenticated attackers read or write database tables. Lovable disputes the entry, because each customer is responsible for protecting the data of their application. Either way, the check is yours to make.

    Source: National Vulnerability Database, CVE-2025-48757

  • Keys and environments

    No secret in the browser or in the repository, and a place to test changes that is not the live product.

  • What happens when it fails

    A declined card, a lost connection, an empty account, an import with bad rows. Each needs a screen and a way back.

  • Money

    Payments tested end to end: a refund, a failed renewal, the same event arriving twice.

  • Knowing when it breaks

    Error tracking, logs you can search, and a backup you have restored at least once.

What to keep and what to rebuild.

  • Keep

    The idea, the data model if it holds, and every screen your users already understand.

  • Rebuild

    The flow the product is bought for, end to end, with every state, on components that agree with each other.

  • Replace

    Anything that touches sign-in, permissions or payments and was never read by a person.

A full rewrite is rarely the right first step. It trades a product that works for a plan.

What it costs in the United States.

Nobody publishes a price list for this work yet, so the honest figures are the hourly ones, and the arithmetic you can do with them.

Figures as published on the pages linked, read on 18 November 2026. Clutch states no sample size, so treat them as direction, not as a quote.

Keep using the tools afterwards.

Going to production does not mean giving up Lovable or Cursor. It means giving them rules.

  • Components first

    A small set of components the tools have to use, so a new screen is assembled from them and not invented.

  • Rules in the repository

    Short written guidance on naming, states and structure, in a file the coding agent reads.

  • A person reads what ships

    Every change to sign-in, permissions and payments is reviewed by someone before it merges.

How a Product Rebuild works

Questions

Before you launch.

  • For a demo or a small private group, often yes. Before strangers sign up and pay, check who can read whose data, what happens when something fails, and how you will know when it breaks.

  • Rarely all of it. We rebuild the flow that matters inside your codebase, frontend and backend, and the rest of the app stays as it is unless we agree otherwise in writing.

  • It depends on how much of it people will use on day one. With Prodular, one key flow redesigned and built full stack is from $20,000, in 4 to 6 weeks. $20,000 covers one key flow in a live product. More flows, or a large new backend, are priced in the written scope.

  • Yes. We work in your repository and the stack you have, say on the first call what the product needs before real users, and put it in the written scope.

  • Yes. The components and rules are documented so Cursor, Claude or v0 build on them instead of inventing a new button every time.

Next step

Show us what you built.

You get a written scope, with the price and the date, before anything starts.