Services / Software Project Rescue
Project stalled, broken or abandoned? Get an honest answer first.
Before anyone writes more code, you get an honest answer: keep it, fix it, rebuild part of it, or stop.
Does this sound like you?
This is usually where the rescue conversation starts.
- “My developer stopped answering, and I’m not sure I still have the passwords.”
- “The agency kept saying ‘nearly done.’ It still doesn’t actually work.”
- “It worked fine in the demo. Now it breaks every time someone real uses it.”
- “Nobody, including the last person who touched it, can explain how this thing works.”
- “I don’t want to be sold a rebuild I don’t actually need.”
- “We’re close to launch, and one problem keeps stopping us.”
What I’d do.
Send me what you have.
Any access you already hold, screenshots, invoices, and a plain account of what stopped working. No judgement either way.
Get a written assessment.
I check the access, the code, the data and the risk, then write it down plainly: what’s actually there, what’s actually wrong, and what your real options are.
Choose the route, on the evidence.
Keep it, fix it, rebuild part of it, or stop spending, based on evidence, not on what pays me the most.
Proof.
A slow site facing security threats, stabilised and secured.
The site was on an old server that blocked its own malware scans and was running outdated plugins. I audited the security, brought in real protection, moved it to a stronger server and got it fast and safe again.
See the iLearn project →Recovering data everyone thought was gone.
After Buffet Olives lost content inside WordPress, I pulled the raw data out of the database directly, rebuilt it, and handed it back as a browsable page and a spreadsheet, so nothing further had to depend on the broken site.
See the Buffet Olives recovery →“Anton was able to help solve a problem which had confounded other WordPress experts. His communication was excellent and he sorted it out quickly and professionally.”
Rachel Case
How it starts, and what it costs.
You’re never committed to more than one piece.
- A free conversation.Send me what you have: any access you hold, screenshots, invoices, and a plain account of what stopped working.
- The written plan.For a rescue, the plan starts as an assessment: what’s actually there, what’s genuinely wrong, and your options: keep it, fix it, rebuild part of it, or stop. Fixed fee, agreed in writing before it starts, and credited against the first piece. The plan is yours to keep.
- One working piece, at a fixed price.What I quote is what you pay for the agreed scope. Every piece after it is quoted before it starts, and you can stop after any piece and keep what’s built.
I can’t promise your project can be saved. I can promise a clear assessment, a plain list of risks, and an honest recommendation.
Nothing beyond the assessment is booked until you decide what happens next, and it’s scoped once we both know what’s actually there.
Not sure where to start? Answer seven short questions →For your technical person
What the assessment covers
- Access audit: confirm who controls the repository, hosting, database, domain, email, payment accounts, analytics and third-party services. No rescue works without access.
- Codebase review: structure, dependencies, database model, authentication, deployment path, and obvious security or maintainability risks.
- Risk register: missing credentials, fragile code, no tests, abandoned packages, unclear data ownership, or a deploy process nobody understands.
- Decision plan: continue, fork, rebuild selected pieces, or rebuild from scratch. The answer follows the evidence, not developer ego.
What rescue work usually uncovers
- Ownership: the domain, hosting, repository or payment accounts were opened by someone else. That gets fixed first.
- Architecture: a project can look almost complete while still missing payment, authentication, admin workflows, deployment and error handling.
- Scope: the project never had a shared definition of done. A written rescue plan turns vague frustration into a clear list of decisions, fixes and tradeoffs.
Honest rescue rule
If the code is unsafe, unmaintainable, or cheaper to replace than continue, I will say so. Rescue means recovering the business outcome, not defending bad code because someone already paid for it.
Cost and guarantees
The first step is the written plan, which for a rescue is a fixed-fee assessment, agreed in writing before it starts. After that, the work is quoted one piece at a time, depending on what the assessment finds.
I can’t guarantee a rescue will ship, because it starts with unknowns. I can guarantee a clear technical assessment, a risk register and an honest recommendation. If the code or hosting can’t be reached legally, recovering the accounts comes first. A mutual NDA is available before you share anything.
What I need to start the assessment
- Access: repository, hosting, database, domain and relevant third-party accounts you legally control.
- History: the agreed scope, invoices, screenshots and a plain account of what stopped working.
- Immediate risk: the deadline, revenue, security, data or operational issue that needs protection first.
Background
26 years in technology, 15 years of that building production software, and over ten years (2002 to 2013) running my own IT businesses before that, so I read an inherited project like an owner as well as a developer. 170+ Codeable projects between 2014 and 2018, much of it exactly this kind of work: picking up and continuing what another developer had already started.
- 26years in technology
- 15+years building production software
- 10years running his own IT businesses
- 170+Codeable projects, 2014 to 2018
Coverage
Based in Cape Town, South Africa. Rescue engagements are also handled for businesses in the United States, the United Kingdom and Australia.
Specialist terms and job titles
Also described as: software project rescue, software recovery developer, legacy codebase takeover, abandoned project rescue, stalled software project consultant, WordPress and WooCommerce recovery specialist. Areas of work: access recovery, codebase audit, risk assessment, WordPress and WooCommerce data recovery, security hardening, hosting migration, rebuild-versus-continue assessment.
More background
- The rescue guide, the plain steps a rescue usually follows.
- About Anton, the full story.
- Full project portfolio.
- Client reviews.
- Blog.
- General site FAQ.
Questions owners ask.
How much does a rescue cost?
It starts with a written assessment at a fixed fee, agreed in writing before it starts. What happens after that depends on what we find, and nothing further is charged until you agree to it. You’ll know the price of each piece before it starts. See how a project with me works →
What if the previous developer won’t hand over the project?
Then the first job is recovering your accounts and access. If something genuinely can’t be accessed legally, that changes the plan, and I’ll tell you early.
Will you sign an NDA?
Yes. Rescue work often means looking at sensitive business and technical information, and I keep that in confidence.
Can you guarantee the project will ship?
No, and nobody honestly can promise that this early. What I can guarantee is a clear technical assessment, a plain list of risks, and an honest recommendation.
We built this ourselves using AI. Does that change anything?
Not much. I look at what’s actually there, the same way I would for any other project, and tell you honestly whether to keep it, fix it, rebuild part of it, or stop. If it’s still early and heading toward real customers, the AI-prototype readiness check may be the faster starting point.
Related help.
Tell me what’s stuck. You’ll get an honest answer before anyone writes more code.
+27 87 150 9305Monday to Friday, 09:00 to 20:00 SAST. Or email info@antondevilliers.com.