Use Cases
Fifteen years. Close to 500 projects. Almost none of them we are allowed to name. Here is everything we can tell you.
Since 9 February 2011
The work is everywhere. Our name is nowhere.
We started Codify on 9 February 2011, back when “digital transformation” was still just called getting the thing to work. In the fifteen years since, we have delivered close to 500 projects — billing counters and booking engines, factory dashboards and field apps, storefronts, schedulers, internal tools nobody outside the company will ever see.
You will not find a logo wall on this page. The overwhelming majority of that work shipped under NDA, or white-labelled — built by us, launched under someone else’s brand, supported by their team, credited to their roadmap. That is the arrangement, and we take it seriously enough that we will not bend it for a marketing page.
So instead of names, here are the patterns: the problems businesses keep arriving with, what we actually build, and what changes on the other side.
~500
Projects delivered
Across retail, manufacturing, healthcare, education, logistics and services.
15 yrs
Of compounding practice
Every stack we have outgrown taught us something the next client did not have to pay for.
Most
Under NDA or white-label
Which is why this page describes problems and outcomes, never clients.
Six reasons people call us
Different industries, different words, the same handful of underlying problems.
“It works, until it’s busy”
The system that was fine at 200 orders a day falls over at 2,000. We find the real bottleneck — usually three queries and one synchronous call — and rebuild that, not the whole product.
“We do this by hand every month”
Someone is exporting a spreadsheet, re-keying it somewhere else and emailing a PDF. That is not a people problem, it is a missing integration. We remove the copy-paste, keep the audit trail.
“We need a product, not a project”
White-label builds where our engineering ships under your brand. You own the roadmap, the customers and the credit. We stay the invisible engine room — and stay under NDA about it.
“Nobody left here understands it”
A decade-old system still running the business, written by people who left years ago. We map it, wrap it, then move it forward in slices — without the big-bang rewrite that takes companies down.
“Can AI do this part?”
Often, yes — the reading, sorting, drafting and matching. We put agents on the tedious middle of a process and leave the judgement calls with your people, where they belong.
“We have the data, we just can’t see it”
Numbers sitting in four systems that disagree with each other. We reconcile them into one view a manager can open on a phone and act on before the day is over.
Then the clock speed changed
The last two years did more to our delivery timelines than the ten before them.
2011 – 2022
Measured in quarters
Scoping documents. Wireframes reviewed for a fortnight. A first working version somewhere around week ten, by which point the business had already moved. Good work, but the feedback loop was slower than the market it served.
2023 – today
Measured in weeks
A working prototype in the first week, in front of the people who will actually use it. AI handles the scaffolding, the migrations, the test coverage and the first draft of everything — so our engineers spend their hours on the decisions that need a human.
Here is the part that matters, though: AI made us faster, not wiser. Knowing which corner is safe to cut, which edge case will bite at month-end, which “small change” is actually a migration — that came from fifteen years and roughly 500 attempts. The speed is new. The judgement is not, and it is the judgement that keeps the speed from becoming expensive.
Four, with the names filed off
Real engagements, described the only way our agreements allow: by the shape of the problem.
Multi-outlet retail
Nine counters, nine versions of the truth
The problem. Each outlet ran its own billing machine and its own stock register. Head office learned what had sold once a week, by email, after someone reconciled it by hand.
What we built. One cloud catalogue and price list, billing that keeps working when the internet does not, receipts printing to the hardware already on the counter, and stock that moves the moment a sale happens.
What changed. Reconciliation stopped being a weekly ritual. Buying decisions started being made on this morning’s numbers instead of last week’s.
Light manufacturing
The whiteboard that ran the factory
The problem. Job status lived on a whiteboard and in one supervisor’s head. Every customer asking “where is my order?” cost somebody a walk to the floor.
What we built. Barcode scans at each stage, a shop-floor screen that shows the queue, and a customer-facing status page fed by the same events.
What changed. The “where is my order” calls mostly stopped. The supervisor got their afternoons back, and the whiteboard became a whiteboard again.
Professional services
Two hundred inboxes an hour
The problem. A team spent its mornings reading incoming requests, deciding which of six categories each belonged to, and forwarding them on. Necessary, repetitive, and nobody’s idea of a career.
What we built. An AI triage layer that reads, classifies, extracts the key fields and drafts the routine reply — then hands anything unusual, angry or high-value straight to a person with the context attached.
What changed. The queue is handled before lunch instead of after it, and the team now spends its time on the cases that actually needed a human.
White-label SaaS
Their brand, their customers, our engine room
The problem. An established company with a strong brand, a real customer base, and no appetite for building an engineering department from scratch to serve it.
What we built. The whole platform — product, infrastructure, releases and support tooling — shipped under their name, on their domain, at their pace.
What changed. They went to market with a product instead of a hiring plan. You may well have used it. We are not permitted to tell you which one it is.
Why we think this compounds
A good build makes four parties better off
Automation gets sold as a way to spend less on people. That has never been the interesting part, and in our experience it is rarely where the money actually is.
The business
Takes on more volume without taking on proportional cost, and finds out about problems the same day they happen.
The team inside it
Stops doing the work a machine should have been doing, and goes back to the work they were hired for.
Their customers
Get answered faster, billed correctly, and told the truth about where their order is.
And us
We get handed a harder problem next time. Fifteen years in, that is still the part we turn up for.
So — what is slowing you down?
Tell us the part of your operation that still runs on spreadsheets, patience or one very tired person. Odds are we have met it before, in some other industry, wearing a different name.
Get in touch