The role in one line
A designer who codes (or an engineer who designs) to act as the catalyst for the client’s new AI Community of Practice – prototyping AI-powered experiences in working code, and embedding AI into how the design teams themselves work.
What they’ll actually do
- Design interactions for AI-driven products: chat, generative tools, recommendation systems, agentic flows
- Build high-fidelity coded prototypes in React to test concepts with real customers
- Use AI tooling (Figma, Cursor, Kiro, Claude Code, custom plugins) to prototype and iterate at speed
- Build reusable AI pattern components – prompt inputs, chat, data visualisation, agent workflows – aligned to the existing design system
- Work across engineers, architects and data scientists, and bring their workflows into the design practice
- Explain technical trade-offs to non-technical stakeholders
Non-negotiables – screen hard on these
- A portfolio of live, coded prototypes. Not static Figma files, not case-study write-ups. Shipped AI features preferred.
- JavaScript, HTML, modern responsive CSS, React. Hands-on and current.
- Advanced Figma. Variables, auto-layout, prototyping, design systems and tokens.
- MCP experience. Designing AI workflows that connect models to interfaces with context-aware data flow.
- Agent building or AI output evaluation. Practical, with a considered view on where AI falls short.