Skip to content

What it costs to make an AI-built app production-ready in 2026

Hiren Chheta, founder.

With Shreesoftic, making an AI-built app production-ready costs $349 for a two-day triage or $1,500 for a five-day audit, then a fixed-price sprint from $2,500 for one to two weeks or from $6,000 for three to five weeks, with half of the first fee credited back if you proceed within 30 days.

Why does an AI-built app need work before real users arrive?

Tools like Lovable, Bolt, Cursor, Replit and v0 are very good at producing screens that work. They are less good at the parts nobody asks for in a prompt: rules about who can see which data, keys kept off the browser, backups that restore, tests around payments, and a way to undo a bad update.

None of these show in a demo. They show when real users, real money and real data arrive. The cost of fixing them depends on what is actually in your code, which is why every price below starts with someone reading it.

What does the first step cost?

There are two ways to start. Both are fixed price and paid in full upfront.

Two-day triage, $349, two working days. A senior engineer reads your repository and hosting setup and returns the five risks most likely to hurt you first, an indicative range for fixing them, and a recommendation: fix, harden fully, or in rare cases rebuild. Choose this if you want a fast, honest read before spending more. If the triage finds nothing that blocks launch, the triage is free.

Production Readiness Audit, $1,500, five working days. We score twelve domains: architecture, authentication, authorisation, data, security, reliability, testing, performance, AI features, observability, deployment and cost. You get a scored report, a prioritised risk register and a fixed-price proposal to fix what we find. Choose this if payments or sensitive data are already live, or if you need a fixed quote rather than a range.

How much does the fix itself cost?

The fix is a hardening sprint. Its scope comes from the triage or the audit, so you know what is included and what is not before it starts.

Stabilisation sprint, from $2,500, one to two weeks. This closes the findings that block launch, fixed and deployed. It suits an app that is close to ready and has a short list of serious problems.

Production hardening sprint, from $6,000, three to five weeks. This covers the full set: permissions on every table, secrets moved out of the browser and environments separated, database migrations and tested backups, failure handling and safe retries, automated tests around money and data, a build pipeline with one-step rollback, error tracking and cost monitoring. It suits an app that already carries real users, payments or sensitive data.

Both are delivered to production, not left on a staging branch. The word "from" matters: it is the floor, not an estimate. Your number comes out of the triage or audit.

If you also want to move the app off the builder's hosting to a repository, hosting and database you control, we scope and price that from the same triage or audit. It does not require a rebuild.

What does a realistic total look like?

Here are two examples to show how the pieces add up. They are illustrations, not quotes.

For example, an app with a small number of launch blockers might take a $349 triage, then a stabilisation sprint at its $2,500 floor. Half of the triage fee, $174.50, is credited against the sprint, so the total is $2,674.50.

For example, an app already taking payments might take a $1,500 audit, then a production hardening sprint at its $6,000 floor. Half of the audit fee, $750, is credited, so the total is $6,750.

Your app could land above those floors. The next section explains why.

What moves the price up or down?

Five things:

  • How many systems are involved, such as payments, email, a CRM or an AI model provider
  • Whether payments or sensitive data are already live
  • How much of the app is undocumented
  • How fast you need it
  • Whether a person must approve every action the system takes

The triage or audit turns those into a number. We don't guess before that. If the number is more than you want to spend, we reduce the scope rather than the price, so you always know what you are paying for.

How do payments and the credit work?

  • Triage and audits are paid in full upfront.
  • Sprints are 50% on signing and 50% on production deployment. Work over $10,000 is split across three milestones.
  • Half of any triage or audit fee is credited against the sprint that follows, if you proceed within 30 days.
  • We invoice in US dollars. UK clients can be invoiced in pounds at the rate on the invoice date. Upwork clients pay through Upwork escrow.

Hourly billing is for emergencies only. Everything else is agreed before work starts. The full table is on our pricing page.

What does it cost to keep the app working after launch?

Nothing, if you are happy to own it yourself. A hardened app with tests, monitoring and rollback is much easier to keep safe.

If you would rather someone else stayed responsible, managed AI engineering is a monthly retainer with a three-month minimum:

  • Core, from $1,500 a month. Monitoring, fixes, updates and a monthly review.
  • Growth, from $3,500 a month. Everything in Core plus a fixed monthly capacity for new work and priority response.

Retainers are billed monthly in advance, and are available for systems we have built or audited.

Why is there no price before anyone reads the code?

Because unknown AI-generated code is the reason estimates go wrong. Two apps that look the same on screen can be very different underneath. One might have permission rules on every table and a working backup. The other might have neither. Quoting both the same would mean overcharging one of you or surprising the other halfway through.

The triage exists to make that first read cheap and fast. If it shows your app is in better shape than you feared, you have spent $349 to find out.

Where should you start?

If your app was built with Bolt, our page on fixing a Bolt app describes the problems we see most often with that tool and how we approach them. For the full offer, including what is not included, see AI product rescue. If you want a free first look, the self-check takes about a minute and tells you whether it is worth talking to anyone yet.

Questions

Why isn't the audit free?

Because the audit is the work. A free audit is a sales call with a report attached; a paid one is an engineer's honest read, and half of it comes back when you proceed.

Can I get a quote for rescue work without a triage or audit?

No. Quoting unknown AI-generated code without reading it is guessing with your money. The triage gives you an indicative range in two working days; the audit gives you a fixed quote in five.

Do you offer discounts?

No. We reduce scope instead, so you always know what you're paying for.

Do you bill by the hour?

Only for emergencies. Everything else is fixed price with a defined scope, agreed before work starts.

Tell us what's breaking, or what's slow.

We take on a limited number of engagements at a time and reply to every message, including the ones we're not the right fit for.