Evidence

Selected work and research

Research, frameworks and projects across the people making, working with and experiencing AI.

Trust by Design

Trust isn't a branding exercise, and it doesn't appear automatically when technology works. It has to be designed: through transparency, control, explainability, honest communication of uncertainty, human oversight and accessibility, so that people's trust matches what the system can actually do.

Trust by Design is the research foundation under this practice. I developed the framework through embedded research with more than 14 public sector AI implementation teams.

  • ExplainabilityCan people understand what the system did and why?
  • OverrideCan they question, stop or correct it?
  • RecoverabilityCan they get back on track when it goes wrong?
  • AccountabilityIs it clear who is responsible?

EquitAI

An AI agent and automation prototype I designed and built for job applications. It started from the recruitment barriers migrant and refugee women face, and asks what AI should automate, what it should only augment, and where people must stay in control.

Product design work

For product design case studies, including design systems, accessibility to the highest BITV criteria and workshop facilitation, see my product design portfolio.

Each project shows which of the three perspectives it involves and which method stages it covers.