// 03 · make it hold

AI Product Engineering / Prototype Rescue

From promising prototype to software real users can rely on.

None of this means you built the wrong thing. It means you built it fast. Now it needs to hold. We audit what exists, keep what works, and engineer the rest to production standard: security, reliability, evaluation, and a release real users can depend on.

duration: 4–10 weeks
terms: Every engagement is fixed-scope, with a written proposal after a diagnostic conversation.

// fit

Who this is for

Founders and innovation teams with a working prototype that cannot yet support real users.

a fit when

The prototype works in demos but you would not put it in front of paying users yet
You need to know what is fragile before you commit more money to features
Real users, a launch, or a diligence process is on the calendar

not the right call when

There is no prototype yet; a Value Sprint tells you whether to build one

// deliverables

What you get

Architecture audit: a scored report across security, reliability, data, testing, and deployability
Your top risks in plain language, each with its blast radius
A prioritized fix roadmap
Product simplification down to the workflow that matters
UX redesign where the interface fights the work
Evaluation of the AI behavior against defined expectations
Security and reliability remediation
Production release

// process

How it runs

Audit → Stabilize → Harden → Ship → Handoff

// 01
Audit
Score the system across security, reliability, data, testing, and deployability; name the top risks and their blast radius.
// 02
Stabilize
Stop active bleeding: exposed secrets, broken auth, data loss paths. Keep what works.
// 03
Harden
Remediate the prioritized risks; add tests, evaluation, and observability around the core workflow.
// 04
Ship
Release to real users behind proper access controls and monitoring.
// 05
Handoff
Documentation and a walkthrough so your team can maintain it. We work ourselves out of the job.

// before / after

The same feature, prototype vs production

Drag to compare. Left is what ships from a vibe-coded prototype; right is what carries real users.

← prototype · production →

// fetch a quote (prototype)
const res = await fetch(
  'https://api.example.com/quote',
  { headers: { Authorization: 'Bearer sk_live_51H...' } }
);
const data = await res.json();
render(data.price);

// POST /api/quote (server route)
const input = QuoteSchema.parse(req.body);
const res = await fetchWithRetry(
  'https://api.example.com/quote',
  { headers: { Authorization: `Bearer ${process.env.API_KEY}` } }
);
if (!res.ok) throw new ApiError(res.status);
return QuoteSchema.parse(await res.json());

prototypeproduction

⇄

// good to know

Common questions

Will you rewrite everything?

Almost never. A rescue keeps what works and replaces only what is fragile. If we recommend a rewrite, it comes with evidence.

The prototype was built with AI tools. Can you work with it?

Yes, including AI-built prototypes; we work from your exported code or a connected repo. Moving you onto a repository you own is part of stabilizing.

Data security, model choice, IP, deployment, and timelines are covered in the site FAQ.

Discuss this workflow

Every engagement is fixed-scope, with a written proposal after a diagnostic conversation.

Discuss a workflow