Exhibit C·1//Self-directed (opens Self-directed in a new tab)
Interface studies
Self-directed work, kept on file for the same reason as the rest: each one exists to settle a single argument.
- Matter
- Self-directed (opens Self-directed in a new tab) — Interface studies · not client work
- Role
- Self-directed · UI, IA, interaction
- Domain
- Onboarding · consumer · finance · analytics · operations
- Term
- 2024 — 2025
These are not client engagements and they are not on anyone’s roadmap. They are studies — products I wanted to think through, taken far enough to answer one question each and no further.
They are here because a portfolio that only shows commissioned work shows you what someone else decided to pay for. A study shows what I choose to spend a weekend on, and what I think the interesting problem is when nobody is scoping it for me.
All five were redrawn into this file so they read as one hand. The original explorations are on Dribbble; what is filed here is the argument each one was making.
Post the screens.Post the decision, with the screens as evidence.
A shot with no argument is a texture. It shows that you can render a thing, which is the cheapest of the skills involved, and says nothing about whether you should have rendered that thing.
A scheduling product for hospital operations — rota, coverage, compliance and cost. The assign dialog states the eligibility test where the decision is made; reports close the loop with the same numbers the dashboard opens with, over time.
Balance first, then the split, then where it went. Four taps from opening the app to a logged expense — the only measure that matters for a tracker, because a slow one stops being used and then it has no data.
A VPN is checked in a half-second glance, so the connected screen has to answer “am I protected” before anything else. Everything measurable sits under that answer, not in place of it.
This one never reached a build, so what survives is the research. The journey map is where the argument came from; the iteration log is what testing changed, round by round — including the two things it proved wrong.
4
Studies on file
On reflection — The discipline was stopping. Each of these could have grown into a fake product with fake metrics; the useful version stops at the point where the argument is made and leaves the rest unbuilt.
Diagrams on this page are redrawn abstractions of the real work, not client screens.