Repair
It is live and it is hurting: outages, slow pages, wrong data, security worries. We add monitoring, stop the bleeding, then fix the causes rather than the symptoms.
Rescue examplesProject rescue & completion
Someone else started it. We will fix it and finish it.
Software projects go wrong for ordinary reasons: a developer leaves, an agency over-promises, a budget runs out before the last 20%. None of that means the work is wasted. We take over troubled and unfinished projects and get them into production.
Sound familiar?
Signs a project needs outside help
One of these is normal. Three or more, and the project is unlikely to recover on its own.
- The launch date has moved more than twice, and you no longer believe the new one.
- It has been “90% done” for months.
- The developer or agency has gone quiet, or gone altogether.
- Every fix breaks something else.
- It works in the demo but falls over with real users or real data.
- Nobody can explain how to deploy it, or where it is hosted.
- You are not sure you own the code, the domain or the app-store account.
- You have been told the only option is to start again.
Three situations
Repair, finish, or take over
Finish
It was never launched. We establish how far from done it really is, agree a fixed plan to reach launch, keep what is sound, and ship it.
Completion examplesTake over
It works, but the person who built it is gone. We secure your access, document the system, put safety nets in place and become your team.
Takeover exampleThe rescue process
Evidence first, then action
Troubled projects are surrounded by opinions. We start by replacing them with facts — what the code does, what state it is in, and what it would take to make it right — and only then propose a plan.
If something is actively on fire — a site down, data being lost — tell us on first contact. Stabilization can start before the audit is complete.
Secure access
Repositories, hosting, domains, databases, app-store and third-party accounts. If you don't control them, recovering them comes first.
Days 1–3Audit
A senior engineer gets the system running, reads the code and reviews the infrastructure. You get a written report: what is sound, what is risky, what is missing.
1–2 weeks · fixed priceDecide together
Repair, finish or replace — with costs and risks for each, and our recommendation. The report is yours even if you take it elsewhere.
One meetingStabilize
Backups, monitoring, error tracking and the handful of fixes that remove the most pain. Tests go around anything that handles money or customer data.
1–3 weeksFix and finish
Work through the plan in two-week cycles with a demo at the end of each — the same way we run new builds.
As scopedMake it durable
Documentation, automated deployment and a clean repository, so that no single person — including us — is ever a point of failure again.
Throughout
Repair or rewrite?
The question everyone asks
Developers love to recommend rewrites. They are more fun than repairs, and usually worse for the client. Here is how we actually decide.
| Repair is usually right when… | Replacement is worth considering when… | |
|---|---|---|
| The foundation | The framework and language are mainstream and still supported. | It is built on something abandoned, obscure or impossible to hire for. |
| The data | The data model is basically sound, even if the code around it is messy. | The data model is wrong in ways that poison every feature. |
| The problems | Issues are concentrated in a few identifiable areas. | Problems are everywhere, with no sound core to build on. |
| The business | It is live, earning money, and can't be switched off. | It was never launched, so there is nothing to disrupt. |
| The middle path | Most often the answer is neither: keep the system running and replace the worst parts one at a time. It is slower to describe and much safer to do. | |
Examples
What rescue and completion work looks like
Flutter app stuck before store approval
An app that works in the demo but keeps getting rejected — finished, hardened and taken through review.
- Flutter
- Firebase
WooCommerce store rescue
A slow, fragile store with forty plugins and failing checkouts — stabilized, secured and made maintainable.
- WordPress
- PHP
- MySQL
SaaS product abandoned at 70% complete
An agency-built platform that ran out of budget before launch — assessed honestly, then finished.
- Laravel
- PHP
- Vue
- MySQL
Inherited codebase with no documentation
The original developer is gone, nothing is written down, and the business depends on it. We take it over.
- Node.js
- React
- MongoDB
- AWS
Production application that keeps falling over
Outages, slow pages and corrupted records — diagnosed with evidence, stabilized, then fixed at the root.
- Python
- Django
- PostgreSQL
- Redis
Questions
Rescue questions
What do you need from me to get started?
Whatever access you have — repository, hosting, a staging link, the contract with the previous developer — and an hour of your time to explain what the software is supposed to do. If you have no access at all, we start by helping you recover it.
Will you criticise the previous developer?
No. We describe the state of the code factually and concentrate on what to do next. Most troubled projects are the result of pressure and circumstances, not incompetence — and a report written to assign blame is not much use to you.
What if the audit says it can't be saved?
Then you will know, with reasons, for the price of an audit rather than another six months of invoices. We will also tell you what can be carried forward — the data, the designs, the lessons — which is usually more than people expect.
The previous developer won't hand over the code. Now what?
It happens. We can help you work out what you are contractually entitled to, what can be recovered from servers and app stores, and whether rebuilding from the running product is more practical than the dispute.
Do you work with our existing developers?
Gladly. A rescue doesn't have to mean replacing people. We often work alongside an in-house developer or a freelancer who needs senior support.
Find out where your project really stands
A rescue assessment gives you an independent, written answer — and a plan you can act on with us or anyone else.