One person, high task density, AI-supported — the chain without the breaks. UX research, requirements engineering and AI-assisted build in one head: evidence becomes machine-readable requirements the AI builds against — tested behaviour, not a pretty demo.
Pick the shape that fits the moment — a fast build, flexible senior capacity, or bringing AI into your team as a lasting capability.
Research-backed, machine-readable requirements go straight into an AI-built, running prototype — tested with real users, not a slideshow.
I plug in as one expert covering research, requirements and build — high task density, AI-supported. Scaled up when you need it, down when you don't.
I set up the method — Agentic Design System, machine-readable requirements, a CI that checks behaviour and accessibility — and coach your team to run it.
Classically a feature runs through research, concept, design and development — with handovers that lose time and evidence. One person with high task density and AI support covers the same chain, without the breaks in between.
Machine-readable requirements let the AI build the right thing in one pass — instead of something similar next to it.
Every requirement carries its chain back to the research. No lost rationale at the handover.
States, flows, WCAG and acceptance are part of the contract — a CI checks them automatically.
AI speeds up the building, not the deciding. What's good stays a human judgement.
An estimate for the design-and-prototyping phase — exactly the part I compress with AI. Research with real users stays untouched.
Tell me the moment you're in — an idea, a gap, or a capability to build — and I'll say honestly whether and how I can help.