1About 4½ minutes
One leader, four stages
A trust leader
A rise in suspensions at two schools, followed from the first sign to what was learnt, with the people closest to it heard on the way.
Blended AI EducationLeadership. Amplified.
The case study
Where these demos come from: one trust’s work with us, so far.16 slidesShowHide
A working case study · October 2026
Redesigning HR with people and AI
How a growing multi-academy trust is learning how its work really happens, before deciding what AI should do. The trust isn’t named, and the pictures of the apps are the Brightwater copies you can try on this page.
Short demos of AI agents helping leaders, and working copies of the apps behind them. They all run on Brightwater, a made-up group of schools: every person, story and number is invented, and nothing you type is saved.
Four demos, about 12 minutes in all.
ListenHow do you find out what is going on?
UnderstandHow do you make sense of it and weigh different views?
DecideHow do you decide, especially when it takes judgement?
LearnDid it work, what did we learn, and who needs to know?
People make the judgement calls, together. The agents gather, check and draft, so the conversation can go further.
Learning that travels. The reasons behind each decision are kept, so the next leader can build on them.
2About 2½ minutes
The grey areas of HR policy
An HR adviser and a headteacher
A request for time off for a hospital appointment: what the policy settles, what it leaves to judgement, and an agreed starting point.
3About 2 minutes
School visits
A leader on a school visit
What a leader sees and hears on a visit becomes a plan, a test, and learning for the whole trust.
4About 2 minutes
Keep-in-touch meetings
The CEO and a headteacher
The CEO and Oakfield’s headteacher prepare for their meeting, and agree what the figures really show.
Play runs a demo with the voice. Present turns the voice off, and you move on with the arrow keys. Escape comes back to this page.
The Clarity story
In depth
One trust leader, from the first sign of a problem to what the whole trust learnt. It works on a phone too.
More demos
Following one request.4 demosShowHide
Our work, as it happens
An agent notices the work while it happens, including the moments where someone uses their judgement.
Every service, from the schools’ perspective
The same case from the school’s side: an agent asks staff what it was like to get help, the way a good interviewer would.
One service, end to end
The request mapped on both sides of every step, with the moments that needed judgement marked.
Why it keeps happening
The loops that bring the same problems back, checked by the people who do the work, and changes tried on the map first.
Try the apps
Working copies of the apps behind the demos, to click through at your own pace.6 appsShowHide
The interviewer
Tell it a real moment from your own work, when it wasn’t obvious what to do. It asks one short question at a time, then gives your story back for you to correct.
Your Voice
Staff sort everyday moments, then tell one story by clicking or talking it through. The results show what everyone said, with the five-person rule for privacy.
The HR space
Our work, Policy check and Sorting what comes in, side by side. See each as an HR adviser, the HR lead or the project lead: their part in the flow of work, what they get back, and what they come to own.
Grounds
An HR adviser’s case, from the first email to a changed policy, in twelve moments. The agent finds the facts that matter, separates what the policies fix from what they leave to judgement, and never makes the call.
Service map
One service end to end, both sides of every step. The real app, with a made-up board: ways in, pictures drawn from the map’s own boxes, and an AI you can ask about it.
Systems map
Why the same problems keep coming back: the loops behind the service map, the people everything goes through, and a change followed all the way round before anyone commits to it.
Brightwater is a made-up trust. These are demo copies: the pages save nothing you click or type. © Blended AI Education