Every decision traces back to evidence
Evidence-based design, in practice: I watch people at work, let them dream the ideal, and log every finding, so each design choice can be traced to its source. AI speeds up the sorting, never the listening.
The chain
Every step of my work leaves a stable ID. Insights name the evidence behind them, and decisions name the insights. Anyone can follow a design choice back to the interview, observation or number it came from.
S
Source
Where it came from: an interview, a field visit, an analytics export.
S001 · About 20 instructor interviews
E
Evidence
What was actually observed or said, without interpretation.
E002 · Weather took five tabs, one location at a time
I
Insight
What it means for the people using the product.
I001 · Instructors think in altitude
D
Decision
What I designed, and which insights it answers.
D001 · One altitude-based weather page
Example from the Instructor Operating Station case study
Services
Three ways to work with me. Each one runs on the same chain of evidence.
UX audit
- Duration
- 8–10 weeks
- Deliverables
- Audit report, Figma file
For a live product that isn't performing. I find what's broken, prove it with users, and hand over a prioritized plan.
Discover
One or two stakeholder interviews on business goals, past experience, KPIs, analytics and existing research.
Evaluate
Expert walkthrough, heuristic evaluation and an accessibility check against WCAG 2.1 AA.
Hypothesize
Name the problems, form hypotheses and prioritize them with RICE or MoSCoW.
Validate
Choose research methods to match the hypotheses, and test them with users.
Synthesize
Triangulate the evidence and rank the issues that held up.
Recommend
Improvements designed in Figma, an implementation roadmap, and the KPIs to track.
New product
- Duration
- 4–6 months
- Deliverables
- Figma file, working prototype, design system
From a blank page to a tested product that developers can build and extend without me.
Discover
Stakeholder interviews on goals, constraints, target users and success metrics; a review of existing data and research.
Research
User interviews and field observation in the real environment, plus a competitor and market scan.
Define
Insights, primary workflows and design principles, prioritized with RICE or MoSCoW.
Ideate
Sessions where users dream the ideal, then concepts, information architecture and key flows.
Prototype
Wireframes, then interactive Figma prototypes, with AI speeding up the iterations.
Validate
Concept validation and usability tests with real users, until the primary workflow works.
Deliver and measure
Final UI and the design system, handoff, launch KPIs, and measurement after launch that feeds new evidence back in.
Redesign
- Duration
- 3–5 months
- Deliverables
- Figma file, working prototype, design system, before/after report
For a product people already rely on. I keep what works, fix what doesn’t, and measure the difference.
Discover
Stakeholder interviews on goals, what’s failing and constraints, and a baseline of analytics, conversion and usage to measure against.
Audit
Expert walkthrough, heuristic evaluation, a WCAG 2.1 AA check, and a funnel and analytics review.
Research and dream
Interviews with current users on what they rely on and what they fight with, plus dream sessions on the ideal version.
Define
Insights, a keep, fix and drop list, design principles and priorities.
Redesign and prototype
New information architecture, flows and Figma prototypes, with the design system built or extended alongside.
Validate
Usability tests comparing the old and new versions on the primary workflows.
Roll out and measure
A phased rollout where it makes sense, handoff, then measurement against the baseline.
Every new product and redesign comes with a design system. Visual in Figma, written in Markdown, so both people and AI tools can build from it.