Approach
The future of product design is not screens. It is judgement systems.
Agents can generate ten versions of an interface before we have decided whether the problem is worth solving. The scarce skill is no longer production. It is knowing what to make, for whom, and why.
How I work
From possibility to impact.
- 01
Make it work
Use AI and prototyping to discover what is possible, early and cheaply.
- 02
Make it good
Apply judgement, evidence and evaluation, then turn that judgement into tools and standards a team can share.
- 03
Make it matter
Bring craft and measurement to deliver value — and learn honestly when a good-looking result hurts the outcome.
The body of work
One thesis, tested from four directions.
Writing, prototypes, patterns and case studies connecting product strategy, recommendations, generative AI and semantic systems.
When agents make the interface
A perspective on the designer's role when producing screens is no longer the limiting step.
Designing the judgement layer
A working exploration of how intent, evidence and constraints shape an interface generated in the moment.
What should the system remember?
Patterns for making memory visible, useful and accountable across AI-native product experiences.
The gap is judgement, not generation
A case study in applying product judgement to recommendations, semantic systems and generative AI.
The premise
Generation makes things. Judgement makes them matter.
- 01
Choose what is worth solving
Start with the decision and the outcome, not the interface.
- 02
Make the reasoning legible
Let people see the evidence, intent and constraints behind an answer.
- 03
Design what the system learns
Treat memory, feedback and adaptation as product materials.
Work with me