ai · prototypes01
Vibe-coded and stuck →
The demo works; production doesn't. We keep what the AI got right and build the system underneath — auth, data, security, scale, support.
Rescue · Safe hands
Broken software?
Bring it here.
The developer vanished mid-build. The vendor spent the budget and shipped a proposal. The AI prototype demos beautifully and falls over in production. The FileMaker system outgrew the person who wrote it. However it broke, someone has to hold it — that's us. We take over what others couldn't, fix what's failing, and when it's time to move, we move the whole solution. Nothing gets abandoned twice.
//. where does it hurt?
Every wreck has
Every wreck has
a way back.
claris · filemaker02
Outgrown your FileMaker? →
Two decades of FileMaker craft, Claris Platinum standard. We support and extend it — and when you're ready, we migrate the whole solution. Not pieces of it.
abandoned · half-built03
The developer's gone →
Vendor collapsed, freelancer vanished, agency moved on. We audit what exists, stabilise it, and take it into safe hands — for the decade after, not the invoice after.
//. how a rescue runs
Triage first.
Triage first.
Then the fix.
A rescue isn't a rewrite with better manners. Most broken systems contain months of salvageable work — real schema, real business logic, real lessons paid for the hard way. The job is to find what's worth keeping, protect it, and give it an owner who won't let go.
Step 01 / 04
Triage.
We read what exists — code, schema, servers, the lot — and tell you within days whether it's a fix, a rebuild or a move. A straight answer, before any commitment.
$ lm triage --patient=yours
Step 02 / 04
Stabilise.
Stop the bleeding first: backups verified, crashes caught, the fragile parts fenced off. The business keeps trading while we work.
$ lm stabilise --uptime=first
Step 03 / 04
Fix — or move it whole.
Some systems want fixing and extending. Some want moving — the whole solution, one rehearsed cut-over, the old system kept warm as the fallback. We recommend what we'd do with our own money.
$ lm plan --bias=honest
Step 04 / 04
Safe hands.
The same senior team on call for the decade after. Rescued systems don't get re-abandoned — that's the whole point.
$ lm serve --since=2006 --ok
//. receipts
We've been rescuing
We've been rescuing
for twenty years.
The $3M failure that started us.
Our first rescue was our founding act: a Fortune 500 trading desk, $3 million and three years down with a global vendor, nothing live. Three months after we arrived, the desk ran on our system — and it still does, two decades later. Read the origin story →
The lab another developer couldn't hold.
TUNRA Bulk Solids' laboratory results and analysis system — sophisticated, scientific, and stranded when its developer moved on. We took it over, stabilised it, and have run and extended it for more than ten years. See the work →
//. straight answers
Rescue,
Rescue,
asked and answered.
Can you take over another developer's system?
Yes — it's one of the most common ways clients arrive. We audit what exists, document what it actually does, stabilise the fragile parts and take over support. The previous developer doesn't need to be reachable or cooperative for that to work.
Our software project failed. Do we have to start again?
Rarely. Most failed projects contain months of salvageable work — real schema, real business logic, real lessons. Triage tells us what's worth keeping, and you only rebuild what's actually broken.
Can you fix an app built with AI?
Yes. Vibe-coded prototypes are usually right about the product and wrong about the engineering. We keep the product thinking and rebuild the foundations — auth, data, security, scale — so what demoed well ships well.
We run on FileMaker and we're hitting its limits. What now?
First we keep it healthy — we're a Claris Platinum partner with two decades of FileMaker work. When it's time to move, we migrate the whole solution in one rehearsed cut-over, using your FileMaker as the blueprint, with the old system kept warm as the fallback.
How fast can you take over?
Triage takes days, not months. If the system is on fire we step in first and let the paperwork catch up — the first job is always to stop the bleeding.