From the science classroom to AI strategy

I started as a physics-trained science teacher and have spent the years since working out how technology can make learning work better for real people.

As a trainee teacher I was selected for Apple's iPads in Schools research programme, and I went on to lead a science department of twelve. A postgraduate certificate in Neuroscience and Education at the University of Bristol changed how I think about design: learning happens in individual brains, and good design respects that.

Since then I've delivered creative learning projects at Learning Pool, supported a £2.9m data programme at the Office for National Statistics, and now lead the learner experience across 56 taught postgraduate programmes and around 3,000 students a year at Learna | Diploma MSc.

There I lead a team of 11, set the two-year AI strategy for educational delivery, built a portfolio of AI tools used across the learner lifecycle, and led a first-of-its-kind MSc from roadmap to launch. Updates across 54 postgraduate courses raised our Clinical Impact Score from 78 to 84.

Every course starts with ADDIE and with the people who will use it: interviews with practising clinicians, conversations with students and direct user testing. Research partnerships with Cardiff University keep the work honest.

I now work with organisations that want the same thing: learning that uses AI well, and can prove it.

Sectors I've worked in

  • Higher education
  • Medical education
  • Corporate L&D
  • Public sector
  • Third sector
  • Schools

How I approach AI in learning

  • Design comes before tooling

    The biggest drop in absenteeism I've achieved came from better course structure, not a new tool. AI earns its place when the design is right.

  • Measurement built in from the start

    Success metrics are set at the design stage and engineered into the tools themselves, so impact is evidenced beyond adoption rates.

  • Build, buy or commission

    Every tool is a judgement about speed, cost and rigour. Sometimes the right answer is in-house, sometimes a partner, sometimes neither.

Applied generative AI

Student support bots, nudging and marking tools built with Claude and the OpenAI API in live delivery.

Learning architecture

Setting the technical direction for how learning is delivered, and what to build versus source.

End-to-end programme design

From strategic roadmap to launch, built to scale from the outset and to meet accreditation standards.

Validation and accreditation

Learning outcomes, assessment strategies and module descriptors written for university validation and Royal College of Physicians accreditation, within OfS and CQC requirements.

Learning analytics

Completion, engagement, pass-rate and feedback measures defined at the design stage and built into the tools themselves.

Digital authoring and platforms

Articulate Storyline and Rise for day-to-day course building, Moodle administration and Canvas implementation.

Assessment design

Assessment strategies for postgraduate programmes, in-house MCQ development and AI-assisted marking.

Training needs analysis

Skills-matrix assessments that show where training is needed and how to deploy people better.

Inclusive by default

Accessible to WCAG 2.1, GDPR-compliant and designed for diverse international audiences.

Programme and project management

PRINCE2 Agile trained, leading concurrent projects in complex, matrixed organisations.

Senior stakeholder influence

AI strategy adopted at C-suite level across a European group of 10+ subsidiaries.