I build AI systems for mission-driven organisations, and teach the
people around them to use AI well. As Senior AI and Technology Analyst at the
Ellen MacArthur Foundation,
I develop the LLM tooling behind the Foundation's research, most recently
a retrieval pipeline that powers a published international policy
analysis, and lead its organisation-wide AI transition: vendor
evaluation, UK GDPR analysis, governance, and structured upskilling
across a 200-person organisation.
Through Intersticia I design
and facilitate AI ethics workshops that get people out from behind their
screens and thinking critically about how these systems actually behave,
including through Brave Conversations.
I started out as a web developer, spending three years shipping
user-facing features in JavaScript and React on the Foundation's public
website before moving into applied AI. I hold an MSc in the Social
Science of the Internet from the University of Oxford, where I studied
how the design of research tools shapes the way people seek and trust
information, questions that now sit at the heart of LLM-mediated
knowledge work.
Selected work
Circular by Nature: AI-assisted policy analysis pipeline
Designed, built, and validated the retrieval-augmented analysis
pipeline behind Circular by Nature (EMF and the Latin America
and Caribbean Circular Economy Coalition, 2026): GROBID PDF parsing,
sentence-transformer embeddings indexed in FAISS, and a hierarchical
roll-up architecture in which the model answers atomic evidence
questions strictly from retrieved context before findings are
synthesised upwards, so every conclusion stays traceable to source
text. Validated against manually identified ground truth on pilot
countries, then scaled to 31 policy instruments across 12 countries.
The methodology is published as a standalone technical appendix.
An embodied workshop in which participants enact the support structure
of a large language model: moving, speaking, and negotiating as the
system they usually only type at. Designed to surface how AI reshapes
group dynamics, it has been run with teams and audiences
internationally, including at WebSci26, and is openly published as a
documented, citable research instrument.
Driving the Foundation's transition to enterprise AI: evaluating vendors
against technical, security, and data protection criteria, authoring
the AI Use Policy and governance framework with in-house legal
counsel, and designing the literacy programme that took colleagues
across every team from awareness to fluency.
