How we work
Honest scoping, tested work, systems you own
A simple four-step process that starts with a free conversation and ends with a system your team can run.
How we work
Four steps, no surprises
01
Free consultation
30 minutes, no cost. You describe the problem; we give you an honest view of its size, its risks and whether we are the right fit. You are in control at all times.
02
Scope and estimate
A written scope with deliverables, timeline and price. Work under 30 days is billed hourly; longer projects can be hourly or a fixed bid.
03
Build and verify
We recommend short cycles so you can measure progress quickly. Every function we write is tested, and AI-assisted work is reviewed by a person before it reaches you.
04
Hand over
Documentation, training for your team, and full ownership of the code and data. Ongoing support if you want it, never required.

Principles
What you can count on
Your data stays yours
We favor private and on-premises setups. We document what, if anything, leaves your network, and you approve it.
Tested and verified
Every function has an automated test, and AI-assisted work is reviewed by a person before you see it.
You own the result
Source code, documentation and data are yours. No proprietary platform and no license fee to keep using it.
Honest scoping
No quote before we understand the problem, and no promised returns. If something is not feasible, we say so early.
Built for handover
Training and plain-English documentation, so your team can run what we build.
Support if you want it
Ongoing help is available and never required.
Data ownership
Why we favor owning your data
We work with cloud vendors all the time, and they are reliable and often the right choice. Where the data is sensitive, or the work has to carry on without the internet, we recommend owning the systems it lives on. Here is the trade-off, both sides.
Privacy
Major cloud providers let you choose the region where your data is stored, and offer strong contract terms. Your data still sits on someone else’s hardware, run by their staff, under their terms and the laws that apply to them. On your own systems, you decide who can reach it.
Access
Cloud services are built for high uptime, but no service is up all of the time, and your own internet link can fail too. A system on your own network keeps working when the internet does not. In exchange, you take on the job of keeping it running, or have us do it.
Price
The cloud costs little to start and you pay as you go, though fees grow with use and the vendor can change its prices. Owning costs more up front, plus upkeep, and then stays steady. Which is cheaper depends on how much you use it; we run the numbers with you. Read our three-year cost comparison.
After handover: a system you own needs looking after. Someone has to apply security updates, update the software and AI models, and test that backups actually restore. Your staff can do it with the written guide we leave, or we can do it for you on a retainer.
The same reasoning applies to AI. Read What You May Not Know About Intellectual Property And AI.
Private by design
Three ways to run it, most private first
Not every organization wants a server in the office. All three options below are real choices, and we help you pick one in the free consultation. Where the data is sensitive, we recommend the first.
Our recommendation
On your own equipment
A server or workstation in your office or data room. Your documents and the AI model stay on your network, and it keeps working when the internet is down. The most private option. You look after it with our written guide, or we do it on a retainer.
A private server you rent and control
A dedicated server at a hosting company, set up for your organization alone. There is no hardware to buy or house. Your data sits in a data center you choose, under a contract you sign.
A business cloud AI plan, with a policy
A business or enterprise plan from a major AI vendor, with settings that keep your data out of model training, plus written rules on what staff may put into it. The quickest and cheapest way to start. The vendor holds your data under its terms.
Built from open, standard parts
We build on open-source software and open-weight AI models, so you are not tied to one vendor and another firm could take over the work.
- Linux servers and standard databases such as PostgreSQL
- Open-weight AI models such as gpt-oss, Llama and Mistral
- Your documents kept in formats you can open without us
Security basics we set up with you
- Encrypted disks and encrypted backups
- A separate sign-in for each person, with access matched to their role
- A record of who used the system and when, kept on your side
- Security updates on a schedule, and backups tested by restoring them
Questions? Read our frequently asked questions.
Find out what your project really takes.
Book a free 30-minute consultation. You will leave with a realistic sense of size, cost and risk, whether or not we work together.