ITSUA
ACCEPTING NEW WORK
§ SERVICES — SOFTWARE PROJECT RESCUE

Software Project Rescue & Recovery

Someone has to take the wheel of a project that stopped moving. We read the code and the way the work is run, say plainly what is wrong, stabilise what is already in production, and then ship. Every rescue starts with a two-day audit — including the version where the answer is that you don't need us.

§ AUDIT FIRST
2 days
We don't quote a rescue we haven't read

No recovery starts with a proposal. It starts with two days in the code and the process, and a written verdict — what is actually wrong, what to fix first, and a date you can plan against.

§ 01

Who this is for

Six situations. Most projects that reach us are running at least two of them at once.

The date keeps moving Every sprint ends with a new date, and nobody can explain what happened to the last one.
Production is unstable Releases break things, the same incidents come back, and the team firefights instead of building.
You inherited the code An acquisition, a founder exit, or an agency that stopped answering. Nobody left knows why it works.
The vendor is going or gone Velocity dropped, quality slipped, or the contract is ending and the handover is a zip file.
Someone said rewrite It might be the right call. It is usually the most expensive way to find out.
It has to be stable before it can grow The roadmap is blocked until the thing underneath it stops moving.
§ 02

What actually goes wrong

Rarely one dramatic thing. Usually five small ones compounding since month three.

Debt nobody priced Shortcuts taken under deadline, never written down, now charging interest on every change.
No spec, so no done Scope lives in conversations. Features get built twice and signed off never.
One person holds the system The build works because someone remembers a step. When they're out, the release waits.
No safe way to ship No staging, no tests, no rollback. Releasing is a gamble, so releases get rarer and bigger.
Architecture that outgrew its assumptions Built for a prototype's load and a prototype's data model. Both changed; the design didn't.
Sometimes it isn't the code The codebase is fine and the way the work is run isn't. We say that out loud when it's true.
§ 03

What a rescue involves

In order. Nobody gets to argue about the roadmap while releases are still a gamble.

Audit Two days across the code, the commit history, the tickets and the people. One written verdict, forwardable as-is.
Stabilise A 60-day fixed-scope phase. The release path, the recurring incidents and the data first — features after.
Take over delivery Our team picks up the roadmap on your process or ours. Weekly demos from week one, staging from day one.
Refactor, or rewrite deliberately Replace the parts that earn it, keep the parts that don't. A rewrite is a decision with a number attached, not a mood.
Replatform and migrate Off a stack you can't hire for and onto one you can. We consolidated a 50-site estate off WordPress, Wix and Squarespace onto a single stack.
Operate it afterwards CI/CD, containerised deploys, observability and on-call playbooks — or a documented handover to your team. One platform we took on in 2016 is still ours to run.
§ 04

Projects we took on

Inherited estates, parallel systems, and platforms we still run years later.

ALL CASE STUDIES →
§ 05

How a rescue runs

01 Send access The repo or a staging URL, plus whoever remembers the history. ↳ ACCESS
02 Rescue Audit Two days reading the code, the last three months of commits, and how the work is actually run. ↳ 2 DAYS
03 Written verdict What's wrong, what to fix first, an honest date — and whether we'd take the work, including no. ↳ VERDICT
04 Stabilise 60 days, fixed scope. Stop the bleeding before anyone argues about features. ↳ 60 DAYS
05 Ship again Fixed-scope phases with weekly demos until the roadmap moves at a speed you recognise. ↳ SHIP
06 Hand back, or keep running Documented handover to your team, a monthly operate retainer, or both. Your call, not ours. ↳ YOURS
§ 06 — TESTIMONIAL
Development Services for Recruitment Agency
“They are always available and are very committed.”
So far, ITSUA has been meeting the client's expectations. They are using Jira for both project management and communication. Their availability and passion for completing the client's goals have been notable.
Anonymous
Anonymous
CEO, RECRUITMENT AGENCY
READ FULL REVIEW ON CLUTCH
§ 07

Software project recovery, answered

Taking over a build that has stalled, broken, or outgrown the team running it — reading it honestly, stabilising what is already in production, and getting delivery moving again. It is the same engineering as a new build, done with less room to guess and someone else's decisions already in the ground.

Yes — most recoveries arrive that way. We need repository access, whatever documentation exists, and one conversation with someone who remembers the history. The previous team's cooperation helps and is not required.

The audit is two days. Stabilisation is a 60-day fixed-scope phase. What comes after depends on what you want shipped — and the audit gives you that date before you commit to any of it.

Usually rescue. A rewrite restarts the clock on every bug you have already found and paid for. We recommend replacing a system when the stack can't be hired for, when the data model is wrong at the root, or when the cost of a change has stopped being predictable — and we bring a number, not an opinion.

Rescue & rewrite engagements run $20K–$80K, quoted as fixed-scope phases once the audit has read the actual code. The audit itself is small and deliberately separate: we don't price work we haven't looked at.

Then that is what the verdict says. The audit covers whether we would take the work at all, and no is an answer we have given before. Some projects don't need an agency — they need one decision.

No. The fastest recoveries keep the people who know the domain. We embed senior engineers into existing teams as a matter of course, and the audit says plainly when the problem is capacity rather than capability.

TypeScript, Node.js, PHP and Python. React, Vue, Nest, Laravel, Symfony, Astro. MySQL, PostgreSQL, Elasticsearch. Shopify Plus, Strapi, WordPress. Docker, Cloudflare, AWS, Azure. We inherit code we would not have written — that is what the job is.

More on pricing, process, IP and where we're based in the studio FAQ.

§ 08 — RESCUE AUDIT

Send access.
Get a verdict.

Two days with senior engineers, one written answer, and no obligation to hire us after it.

BOOK A RESCUE AUDIT
EMAIL hello@itsua.com HQ Sakala tn 7-2, Tallinn, 10141, Estonia VAT EE102549430 TIMEZONE EET / UTC+2 RESPONSE SLA < 24h