Why are you growing with growing pains?Grow without growing.
The old rule was that doubling your revenue meant doubling your structure: more people, more space, more process. Hybrid working only changed where people sit, not what they do.
We train a model on your own data, information and knowledge, so work that took weeks or months comes back in seconds, without a line of code, and it never leaves your walls. The only thing you scale is the server where the model lives.
Your people are working harder. But is your business moving faster?
Turn your organisation’s knowledge into scalable capability. Respond to more opportunities without proportionally increasing headcount.
The invisible work
The work that grows fastest is the work somebody has to think about, again and again. That is the work we take on.
Growth used to mean structure, and the costs arrived before the revenue did. Hybrid working took out some of them. It did not take out the repetitive cognitive work.
It never appears as a line in your budget, because it is spread across the people you already employ. And not all repetitive work is the same.
Repetitive work
The same motion again and again, with little judgement. Scanning, filing, copying, moving files from one place to another.
Repetitive cognitive work
The same judgement again and again. Reading, checking, matching, classifying and deciding what needs attention, like matching an invoice to its purchase order. This is what you hire skilled people for, and it is what we take on.
We focus on the second kind, and it is expensive because of who does it: the people you hired for their judgement spend their week on the part of it that repeats. Take that off them and they get their time back for the thinking that does not repeat.
Where do we find repetitive cognitive tasks?
- Something arrives, or it is already on someone’s list for today
- Read it, pull out what matters
- You record it in your system
- Check it against your rules, your records and what your people know
- Draft what goes back out
- A senior person reviews it and gives feedback
- Your team approves and sends
The repetitive cognitive tasks — what we train the model on Your people and your systems, unchanged
There is no code in this. You are training and tuning a model to use your data, your information and your knowledge, which is why you are not back with a developer every time something changes.
Growing no longer means expanding your org chart. It means expanding your architecture.
Why now
Four lines on your P&L. Every one of them starts moving the right way the day your first set of tasks goes live.
Revenue you are turning away
Whatever wins work in your business — tenders, proposals, quotes, applications — the ones you miss this year are gone, not deferred.
Cost that stops following revenue
More volume stops meaning more people, more space, more process.
Cash in sooner
The document that triggers the money goes out in days, not weeks.
An asset, not a subscription
No per-seat licences, no per-token fees. What you spend buys something you keep.
What we do
We take the repetitive cognitive tasks inside one of your processes and train a model on your own data, information and knowledge to do them. Four things hold, whatever the work.
One set of tasks, then the next
One set of repetitive cognitive tasks inside a process. Then the next set, when you are ready.
On your own hardware
Runs on-premise, inside your own walls. Or in your private cloud, if you prefer that.
You own the model
Your data, your information, your knowledge. No usage-based token fees.
Governance from the start
The right ownership and governance agreed before anything runs: who decides what, and what stops for a person.
A worked example
Tendering for a construction or engineering contract. An illustration, not a client case, and nothing in it is unusual.
You set the rules. What goes straight through and what stops for a person is your decision, task by task, and you can change it. Nothing runs on its own because we thought it should.
This is only possible because of what you already have: the proposals your people have written, what you know about your own projects, your subcontractors, your rates, your people. Work that has been checked many times over and that you trust. The judgement stays with them. The tedious, repetitive part of assembling it does not.
And it compounds. Every correction a senior person makes goes back in, so the next draft starts closer to the mark than the last one did. It is not a document generator that performs the same on day two hundred as on day one.
Private AI, and why it has to be private
Private means two things, and you need both. The model learns from your documents, so what comes back reads like your firm rather than like everyone’s. And it stays on your hardware: your own IP, and everything you hold about your customers, your suppliers and your people, stays inside your walls.
Training it on your work is what makes the result yours rather than generic. Keeping it on your hardware is what keeps the cost flat while the volume climbs.
How we work together
Scoping comes before the project and is free. The project itself is two stages. Running and supporting it comes after.
-
1Scoping
Free. An orientation call, then we agree together which use case to start with. You have the scope, the requirements and a fixed price before the project begins.
-
2Map and govern
We map the repetitive cognitive tasks and agree the business requirements, before any training or tuning.
-
3Train, tune and connect
The model learns your documents, your language and how your people decide. We connect what has to be connected: the triggers that start the work, and where the outputs go.
-
4Run and support
You go live, with a warranty period. Ongoing support is optional and quoted separately.
- High-level scoping comes first, and is free
- The project: typically 8 to 12 weeks, at a fixed price
- Run and support comes after, quoted separately
Clear governance throughout the project. We report continuously on progress, risks and the decisions needed, so we move at pace together.
Critical success factors
- Your systems set the timeline, not the work
- Business, technology and risk, in from the first call
- Take the tasks as they run today
- A business case you go back to
Where it fits
Functional areas. Think of accounts payable, procurement, HR, contracts and legal, forms, applications and grants, safety records and permits, leases and property, compliance records — everything you have in there.
And in any industry. Real estate, higher education, legal and professional services, mining, logistics and transport, engineering and construction.
Of course there are things specific to an industry. We do not have to be the expert in yours — you already are, and the model is trained on what you know.
Your best people are buried in work that could be done differently.
Unlock your organisation’s knowledge. Reduce repetitive cognitive work. Pursue more opportunities with the team you already have.
Who you work with
Doing this well takes business knowledge, technology, an understanding of how a process actually runs, and the judgement to make decisions inside a real organisation. Between the three of us, we bring all of it.
Business and delivery · Australia
Marcela Ugaz
National lead partner for mid-market across two Big 4 firms, in Australia and Europe, running the practice as well as the engagements. Business and technology transformation, and enterprise operating model design. Based in Sydney.
Technology · Peru
Ernesto Cuadros-Vargas
More than twenty years in computer science. PhD from the University of São Paulo, developed with Carnegie Mellon and TU Berlin. Has served on the Board of Governors of the IEEE Computer Society.
Strategy and operations · Peru
Max Ugaz
Four decades in technology and management transformation. PhD in information systems engineering, an MBA, and a graduate business school professor. Member of the Peruvian AI Chamber.
Photographs to come.
Growing Without Growing works locally. We scope, contract, manage and support your project in Australia, with a business project manager and account manager based here, on Australian time. Our engineers in Peru — engineers, not developers: they train and tune the models for you, in your infrastructure. They work to your security protocols, and your documents stay on your premises.
Questions
The ones buyers ask first. We already know them, so your scoping call can go to what is specific to you.
What does it cost?
It depends on the process, so we start with the scoping call. For the right process, payback could be as little as 2 months, and we estimate it with your own numbers.
What changes for my team?
They stop doing the repetitive part and keep the judgement. The model returns a first version; where a call needs experience, it flags it rather than guessing, and a person makes that call. This is built to let you grow without adding headcount at the same rate — it is not a redundancy tool, and the corrections your people make are what keep improving it.
Do we need our own hardware?
Yes. The system runs on your own hardware, inside your own walls. We help you size it during scoping, and hardware is quoted separately.
Do we pay per use?
Not in token fees. The model runs on your hardware and you own it. Ongoing support is optional and quoted separately.
Where does our data go?
The system runs on your premises.
Also on the scoping call: who is accountable if something goes wrong, how your data is handled during the project, and how you audit what it does.
Start with a free scoping call
You pick a time. We talk through how it works, look at the process you want to start with, and give you a first view of the likely return.
Prefer to talk? Call +61 475 043 070