Services
I design products and build systems that ship them.
Lead product designer with 10+ years shaping complex products across fintech, education, AI, marketplaces and media. I take products from ambiguous idea to scaled platform by combining product strategy, agentic systems and craft.
What I do
Deciding what to build
Design strategy
Turning a vague product problem into a few bets you can ship and measure. Research that ends in a finding, not a deck.
- Product bets with a success metric each
- Research synthesis: interviews, tickets, funnel data
- Problem maps and early wireframes
Making it shippable
Design systems
Systems built to be shipped and maintained, not admired. Primitives, parts, applications — solve a problem once, compose it everywhere.
- A component library in three levels
- Interaction states and edge cases designed, not found in QA
- Handoff your engineers can build from
Shipping without the queue
Agentic pipelines
Design velocity capped by frontend capacity is the bottleneck. This closes it: the component library in the repo becomes the source of truth, agents wire it up, and handoff becomes a pull request.
- A written architecture proposal
- Component library in the repo as the source of truth
- Agent roles with a written contract between phases
- A sandbox where stakeholders review working software
Method
How the work actually runs
The same loop, at whatever scale the engagement allows. It repeats per phase — what phase one teaches reframes the problem for phase two.
Frame the bets
Understand the business, its goals and its constraints, then turn them into a small number of explicit bets with success metrics attached.
Research in depth
Customer interviews, support tickets, funnel analytics, behaviour mapping. Synthesised into jobs to be done and rated by severity, so prioritisation is an argument I can show you.
Build the system
Robust visual systems built to be shipped and maintained: primitives, then parts, then applications. High-fidelity where it counts, with edge cases and interaction states designed.
Close the loop
Tight feedback loops with AI, and continuous tuning of the design-to-code pipeline against your organisation's own context and data.
Engagement models
2–3 weeks
Discovery
When you know something's wrong and don't yet agree on what. It ends in a written document — the finding, the bets it implies, a phased plan with a metric per phase — that you can circulate and argue with, whether or not I do the work that follows.
- Customer interviews and funnel review
- Problem map and jobs-to-be-done, rated by severity
- A prioritised set of bets with success metrics
- Low-fidelity wireframes to make the next conversation concrete
4–6 weeks
Pipeline setup
For teams whose design velocity is capped by frontend capacity. I stand up the design-to-deployment pipeline and hand it over documented, so your designers keep shipping after I leave.
- Component library in React and Tailwind
- Agent setup, conventions and context files
- A sandbox environment for testing against real-shaped data
- GitHub workflow, plus a walkthrough with the team
Ongoing
Embedded build
I join the team as a design engineer and stay through shipping — design and production frontend in the same pair of hands, on a weekly release cadence.
- Product design through to production frontend
- A component library maintained in your repo
- Weekly releases and pull-request review
- Working alongside your engineers, not throwing files over a wall