People-first AI for small and mid-sized businesses
Your people, multiplied.
Get more from the team you already have, with AI they trust and a clear view, every month, of what it saves and what it costs.
Everyone is trying AI. Nobody owns it.
Harbour Freight is a fictional company. Not a client's figures.
What we do
Foundations built together.
A portal that makes them last.
Getting real value from AI takes five things: a safe setup, a confident team, the first agents built, a close watch on what it all costs, and a decision each month on what comes next. We build the foundations with your team. The portal keeps them standing.
Set it up safely
Which data AI may use, where it is processed, and who approves what. Agreed before anything is built.
Bring your team with you
Hands on with their own work, so they know what AI is good for and what is safe to put into it.
Build what they need
Their best ideas become working agents, each with an owner and limits from day one.
Keep watch
What is running, who uses it, what it costs, and what the same work cost by hand.
Decide what comes next
Every month: what to keep, improve, train people on or stop. We advise; you decide.
In a smaller business, AI usually sits on someone's desk on top of their day job. Without the time or the specialist knowledge, it waits behind everything else.
We bring the expertise, and we share it. For a week we work beside your team: everyone learns a way of working with AI they can keep using, and the best ideas become the first agents. Then the portal makes it last. It keeps watch every day, and once a month we sit down with you and decide what comes next. You get the judgement and the evidence to back it, without building an AI team of your own.
Where we stand
For your people.
A small team can only do so much in a day. We give each person AI that takes on the routine and amplifies what they are good at, so the same team can take on more.
Their ideas get built
People back what they designed themselves. So the ideas come from your team, and the best of them become working agents.
They stay in charge
Every agent has an owner on your team, and changes to it are theirs to make. We advise; they decide.
They get better at it
Where a team is not using what it has, training comes before cutting. People learn what AI is good for, and what is safe to put into it.
Their time comes back
We count the hours an agent gives back to the people who use it, not only the pounds, and talk about what those hours go to.
nurizon Portal
Three questions, answered every month.
The portal finds the AI agents in your business, measures each one against what the work cost by hand, and turns what it finds into ideas your people decide on.
How is it performing?
Who uses each agent, how well it runs, and where it stands: keep it, improve it, train people on it, or stop it.
What's the ROI?
What the work cost by hand, less what it costs now: the time people still spend and the AI bill. Measured and estimated figures are kept apart, and unknown is never shown as nought.
How can we improve?
What the portal finds becomes a costed idea, checked against your guardrails. The person it belongs to decides, and a month later we check it did what it promised.
In September your agents returned £12,868 after their AI cost, and released 446 hours of work.
Net return by month
After people's time and the AI bill. 2 agents measured, 6 estimated.
Waiting for you
2 decisions this month
Harbour Freight is a fictional company from the portal's guided walkthrough. No figure here is a client's.
It reads usage and cost. Never your prompts, never your outputs, and it never runs anything.
How it starts
Setting you up for success.
A week with your team, then a review every month. The week sets up the guardrails, your people's confidence and the first agents. From then on the portal keeps watch, and we go through it with you.
Before the week
Connect
We agree the guardrails and connect the portal to Claude and Microsoft 365. It reads usage and cost, nothing else.
Days 1 and 2
Workshop
Your people map their own week and get hands on with AI on real tasks. Together we agree what each job takes by hand.
Days 3 to 5
Build
The best ideas become working agents, each with an owner and limits. By the end of the week they are at work.
Every month
Review
How each agent is performing, what the return is, and how to improve, with the decisions waiting for you.
Security and governance
Your data stays yours. Your decisions stay yours.
Security is part of the week from the first conversation, not something added at the end.
What we build
- What we build runs in your own accounts, under your logins and your permissions
- Before the week we agree which data agents may use and where it is processed, including whether it stays in the EU
- Each agent gets only the access its task needs
- Your people leave knowing what is safe to put into an AI tool and what is not
What the portal does
- Reads usage and cost by workspace, key and model. Never prompts, never outputs
- Guardrails you set decide which changes need whose approval. We advise; you decide
- Every decision about an agent is recorded: what changed, and who decided
Who it's for
Built for businesses without an AI team.
You run a small or mid-sized business. Some of your team already use AI, in seats or an API bill or on their own accounts, and someone has asked what it is doing for you. Nobody's job is to answer.
We're paying for this. Is anyone using it?
Which of these agents actually saves my team time?
What is the AI bill buying us, and should it be bigger?
Not for businesses that already have an AI team and their own tooling.
What we build
Your team's ideas, built to last and to pay their way.
Some come out of the workshop. Others come out of a monthly review, when the portal shows something worth building. These are where we expect to start.
Onboarding
Getting a new starter up to speed: the documents, the accounts, the questions everyone asks in their first weeks.
For exampleNew starter guide
Customer service
Answering well without losing the person on the other end: finding the right answer, drafting the reply, summarising the case.
For exampleReply drafting
CRM optimisation
A CRM people trust, without the admin: notes from calls and meetings captured, records kept tidy, next steps written down.
For exampleCall notes to the CRM
Big data analysis
Asking questions of the data you already hold, in plain language, and getting an answer your team can check.
For exampleWeekly report
About
Why we built nurizon.
Between us: deep work with people through change, and leading AI efforts, in organisations large and small.
The tools were there. Again and again, what held AI back was that the foundations for people weren’t in place.
Smaller organisations have a huge need, and huge potential. With fewer layers, they can move fast once their people are on board.
So nurizon starts with the people, and builds the AI around their work.
Why nurizon
New horizon, in one word. Not the far horizon of a strategy deck: the near one, the one your team can reach this week. Said “new-rizon”.
nurizon Weekly
The week in AI at work, every Saturday.
What changed, read for businesses like yours. Every story links to its source.
Read the latest issueGet in touch
One conversation is the place to start.
Tell us about your team and what eats their time.
- info@nurizon.ai
- nurizon
- Based in
- London and Austria