AI that works when nobody is watching.
Most businesses meet AI as a chat window somebody remembers to open. That is the smallest version of it.
It runs in the background.
Not a tool your people have to remember. Work that used to wait for someone to have a free hour happens continuously, on your own systems, and what it produces is waiting when your team arrives.
It serves the whole team, not one person.
One system, used by everyone whose work touches the same material. What one person works out about how to use it, everybody gets. That is the difference between buying licences and having a capability.
It explores what you have no time to explore.
The version of the question you would ask if you had a spare week: the market nobody checked, the seven offers nobody compared, the option that was never on the table. You get the options and the reasoning behind them. You still choose.
It changes what you can sell, not only what you can finish.
Some of what we find makes an existing week shorter. Some of it becomes a service you could not have offered last year. We test both in your own operation, with your own data, before you commit to either.
Three ways to work with us.
-
Consult
We work out what AI changes for you and what it does not, and put it on one page you can act on.
Read more
We learn how you make money and where you are stuck, go through the week with the people who actually do the work, and mark where AI opens something new for you — and where it must stop, because your name, your judgement or your relationships are on it. You get one page, read back to you line by line, and you approve or reject each line before anything is built.
Access to your business and your leadership, in a shape we agree with you first.
-
Enable
We put AI to work in your hands and leave you able to run it without us.
Read more
Written rules for what may and may not be pasted into anything, the tools chosen and configured on your own accounts, your team trained on your own documents, and a named owner on your side for every system who has already changed a rule in it. This is also the Head of AI role when you want it standing: direction, experiments, tool decisions in writing, and the expert in the room when a client or an investor asks about AI.
A place at the table when AI is decided, and access to the people whose work it touches.
-
Build
When the answer is a thing rather than a habit, we build it.
Read more
Systems that run on their own, on your own material, for everyone whose work touches it — quoted per system after we have seen the work, never before. It runs in your accounts, under your logins, with your keys, and the code sits in your repository. Each one ships with a written handover, because the point is that it keeps running whether or not we are still here.
The people who do the work, for the hours it takes to get the detail right.
What the Head of AI role covers.
A running list, in the order it usually comes up.
- Agents that run without anyone starting them: work that used to wait for someone to have a free hour, happening continuously on your own systems
- A new experiment most weeks: a service you could sell, a process you could halve, a tool worth a real test
- Which AI tools to trust with your data and which to refuse, decided for you, in writing
- Client pitches, tenders and grant applications that need AI in them, and need it to be credible
- What your competitors and your field are trying, and which of it is worth copying
- Questions from staff, answered at their desk rather than in a workshop
- The written rules: what AI may and may not do in your business, kept current
- For research groups: which models can do what today, which experiments are worth a PhD student’s month, and where the science must stay human
How the consulting work runs, step by step.
- 01
First: we learn your business.
How you win work, where the money leaks, what your people spend their days on, and what you have quietly stopped believing is possible. Not a workshop, not a questionnaire. We sit in the building and ask questions.
- 02
Then: we map what AI opens, and where it stops.
One page. On one side: the services you could offer, the speed you could reach, the costs you could cut. On the other: what stays with your people, because your name, your judgement or your relationships are on it. Read back to you, and you approve or reject each line before anything is built. That page is the plan.
- 03
Then: we prove the first ideas live.
In your office, with your data, in your own accounts. Two or three ideas from the page get tested. The ones that work stay. The ones that fail cost you an afternoon and teach you something.
- 04
Last: your team takes over, and we decide what is next.
The people who will run what works learn to change the rules without us, and each system gets a named owner on your side who has already changed one. Then we look at the page together and decide whether we take on the Head of AI role for the months ahead.
How it is priced and how it ends.
- A fixed price for the engagement, quoted in writing after the free first call. One invoice, no day-rate meter, no expenses.
- The Head of AI role is monthly, and you can stop with one month’s notice, any month, without giving a reason.
- Model and cloud usage is billed to your own accounts, at cost. We never resell it and we never mark it up.
- Building the larger systems the engagement finds is quoted separately, per system, after we have seen the work. We never quote a build before that.
- If we conclude early that there is nothing here worth the money, we say so, we stop, and we bill only for the days already spent.
What stays with your team.
- Every decision
- Every signature, grade, valuation and diagnosis
- Every client, patient, student and colleague relationship
- The choice to stop any month
A field we have never seen.
A manufacturer, a logistics group, a private clinic, a fund. What we build changes completely from field to field. How we find out has not had to — and if your field turns out to be the exception, you hear it early rather than at the end.
A small company, on purpose — and nothing that stops with us.
Dinobridge is a small company built around one senior AI expert, by design. You get the person who built the systems, at your table, with no account manager between you and the work and no junior learning on your data. That is only safe if nothing important depends on us — so it does not.
Everything is yours from day one.
Every system runs in your accounts, under your logins, on your infrastructure. The code sits in your repository. The API keys are yours and billed to you directly, so you can see and cap what they cost. We hold no credential you cannot revoke in an afternoon.
Every system gets a written handover.
Written the week the system is built, not the week we leave: what runs where, which service does what, how to change the rules, how to switch it off, what breaks if you do, and what to look for when something looks wrong. Written for the person on your team who will own it, and tested by having them use it while we watch.
A named person on your side owns each system.
We do not hand over to “the team”. Before anything goes live, one of your people is trained on it, has changed a rule themselves, and is named in the handover as its owner.
Response times, in writing.
Same working day for anything that is broken, two working days for everything else, and a named window each week when you know where to find us. If we are away for more than a week, you get the dates in advance.
If we stop, nothing stops.
No system needs a licence, account, server or subscription of ours. If we disappeared tomorrow, everything we built would keep running and your named owners could keep changing it. That is the test we design against, and you can put it in the contract.
We work on a mandate, not as staff.
The Head of AI role is a mandate with its own deliverables, its own tools and several clients — not an employment relationship, not instruction-bound work, and not integration into your organisation. If your finance or HR function wants that confirmed in advance, we will say so in the contract and we are happy to have the status formally determined first.
Where we are not the right supplier.
- No 24/7 on-call rota, and we do not hold your production systems. If you need an operations team with a pager, you need a managed-service provider.
- No ISO 27001, SOC 2 or TISAX, no security team, no penetration-testing programme. If your procurement requires certification, you hear it before you spend time on us.
- A small number of clients at a time. When we are full, we say so.
- Nothing that takes a decision, a diagnosis, a valuation, a grade or a signature away from a person. We have turned down work over it, and we would turn down yours.
What AI already does inside teams like yours.
Every example here is one we designed and ran ourselves, in a real business with real deadlines. Pick the one closest to your work.
A sector we have never worked in? The work runs the same way, and you hear early if it does not. Ask us.
Tell us what you are dealing with.
Thirty minutes, free, no preparation. We tell you honestly where AI could change something for you, and where it would not.