Apply
A new kind of school · 2026 intake open

We don't do GCSEs. We don't do A-levels. We don't believe a timed written exam tells you anything useful about a fifteen-year-old. At Catalyst, students aged 11–18 build, ship and defend real AI projects — and leave with a portfolio that proves what they can actually do.

Apply for September → See the programme
Scroll
Where we stand

We walked away from exams because exams don't measure the right thing any more.

A two-hour written paper, sat in silence, with no tools, on a topic you'll never revisit — that is not what working life looks like. It hasn't been for years. The students who thrive after school are the ones who can scope a problem, gather what they need, ship something that works, and explain it under questioning. So that's what we teach. That's what we assess. That's all we do.

— The reason Catalyst exists.

The programme

One school. One qualification. Seven years of real work.

Every student at Catalyst is on the same path: the Catalyst AI Diploma. From Year 7 to Year 13, they work in increasingly ambitious project cycles — building, presenting, defending, and adding to a portfolio that follows them out of the school gates and into university applications, apprenticeships, and their first companies. Faculty use AI as a teaching instrument: tutoring, marking, simulation, retrieval practice. Students use AI as a tool: a co-engineer, a research assistant, a critic.

01

1:1 AI tutoring

Every student has an AI study partner that knows their projects, their reading, and where they got stuck last week.

02

Build & defend

Every project ends in a viva. If a student can't explain it, they didn't make it. Provenance is recorded; honest AI use is graded.

03

Real artefacts

Code, hardware, models, papers, films, prototypes — every student leaves with a body of work, not a slip of paper.

04

AI literacy spine

Prompting, evaluation, bias, provenance, responsible use — taught explicitly from Year 9 and applied from then on.

The story

It started with a robot.

In the lead-up to the 2021 semi-finals of the ANA Avatar XPRIZE — a global competition with a $10M prize pool for fully teleoperated avatar systems — our students entered. The serious field was university labs and industrial research groups.

We qualified. The University of Sheffield qualified. We were the only two teams from England that made it through. COVID stopped us from going further.

But by then we knew. These students hadn't reached a global semi-final by being drilled for exams. They reached it by building, breaking, fixing and defending a real piece of engineering for two years. Catalyst is the school we built so the next group could do that from the first day.

What the students built

  • PlatformFully teleoperated avatar
  • VisionOpenCV on Raspberry Pi with Intel Neural Compute Stick
  • HapticsCustom system, designed in-house
  • FrameAluminium, machined on-site
  • PowerIn-house Li-ion battery packs
  • Stage reachedXPRIZE semi-finals (2021)
At a glance

The numbers.

0Max class size
0Timed written exams
0Diploma projects per phase
0Nationalities on roll
"Our job isn't to compete with AI. It's to graduate students who can direct it, audit it, and outthink it where it matters." — Head of School

The students who build with AI now will define what comes next.

A polished exam script no longer proves much. A working AI-built artefact, defended in person, proves a great deal. We graduate students who can show you what they made — and explain every choice in it.

Start your application →