Featured
RemoteShe
Remote work, without the trade-offs. A job board where roles are scored by the family-friendly perks that actually matter.
Job BoardCare SignalsShipped
Shipped
Independent products. Solo-built. Across stages.
Research instruments, civic tools, and production platforms, all built around the same question: what gets left out when systems aren't designed for the people doing the most work?
How to read this page
01Stories stay first. Featured platforms and long-form case studies below, grouped by what each product tests.
Flagship
Featured platform
FemHealth Research Finder
Free, open-access women's health research aggregator. One query fans out across four default sources in parallel, with thirteen opt-in sources in Advanced Search, returning sex-equity signals next to every result.
Research DiscoveryWomen's HealthLive PlatformUI Redesign
Featured projects
One research stack. Shipping in production.
Featured
ReturnKit
A planner for the version of you that comes after the first year. For women returning to work after maternity, juggling a career and a household, and anyone whose week does not look like a clean 9-to-5.
Return-to-workPlanningMaternal Intelligence
Featured
MomOps
Why the most complex operational intelligence on Earth is structurally invisible to institutions.
Framework DevelopmentOperational IntelligenceSystem Design
Tools & experiments
Adjacent surfaces, sister tools, prototypes
Archive
Other versions of this work.
Experiments and fragments that do not behave like case studies. They stay here as a living archive, not a publishing schedule.
Archived experiment
Decode
A blood-test reader exploring how AI can explain health data without inventing missing values.
Archived experiment
Plain Stack
Designer-to-stack literacy through journeys, flip cards, and shipping scenarios.
Origin
Universe
A navigable origin system about bodies, mystery, and the first universe.
Field
Creative archive
Notes, experiments, and the wider archive around the work. It does not need to update often to remain meaningful.
Systems behind the work
The recurring failure patterns underneath what I ship.
Shipped work is the proof. Systems thinking is the layer underneath: where maternal, cognitive, psychological, and enterprise constraints repeat across products.
Maternal Systems
Maternal systems fail because care load is invisible to organisational metrics. The most complex operational intelligence on Earth is structurally excluded from every system that measures productivity.
Failure pattern: isolation from the network
Cognitive Systems
Cognitive systems fail when interfaces assume everyone processes information the same way. Numerical exclusion, attention fragmentation, and literacy barriers are treated as user error instead of design failure.
Failure pattern: access degrades by design
Psychological Systems
Psychological systems fail when interfaces re-traumatise the people they claim to help. Trust breaks when information is fragmented, when consent is performative, and when safety is an afterthought.
Failure pattern: path ends with no exit
Enterprise
Case studies from professional practice
Shipping design at scale across life sciences, fintech, and regulated enterprise environments.
CRM Life Science
Making scientific complexity navigable. CRM infrastructure for regulated environments where generic tools fail.
CRMLife ScienceCustomer Relationship Management+1
6 months for initial development and deployment
Payment Devices App
Reducing friction in critical financial flows. Redesigning refund and tip operations used by 30,000+ merchants daily.
FintechMobile BankingUX Research+1
8 months
Design Systems for Life Sciences
Design infrastructure for regulated science. 20+ applications, 100% compliance, 60% faster development.
Design SystemsLife SciencesCompliance+1
14 months
Currently sharpening
Architecture and orchestration, not syntax drills.
As I ship more, the highest-leverage skill layer is no longer typing code. I already work across React, Tailwind, Supabase, Figma, and AI-assisted workflows. What pays now is system design, prompt orchestration, and review discipline. Making sure the output of an agent matches what the product actually needs.
I switch between Cursor and Windsurf, but my preferred combo right now is Windsurf with Claude Code. The GitHub ecosystem sits underneath everything as my deployment baseline.
Where I am sharpening
- Windsurf + Claude Code as the main orchestration combo for production work
- Prompt engineering for production code, not playground tasks
- Spec-first workflows so coverage and tests fall out of the spec, not added later
- GitHub ecosystem (Pages, Actions, Issues) as the deployment baseline
Workbench
The full stack
From Figma to agentic AI. The tools I use to research, design, and ship.
Design & Research
AI & Development
%22%2F%3E%3Cpath%20d%3D%22M14%2038c4-12%2012-18%2018-14s10%204%2018-6%22%20stroke%3D%22%23fff%22%20stroke-width%3D%223.5%22%20fill%3D%22none%22%20stroke-linecap%3D%22round%22%2F%3E%3Cpath%20d%3D%22M14%2044c5-12%2013-17%2019-13s9%203%2017-7%22%20stroke%3D%22%23fff%22%20stroke-width%3D%222%22%20fill%3D%22none%22%20stroke-linecap%3D%22round%22%20opacity%3D%22.5%22%2F%3E%3C%2Fsvg%3E)
Infrastructure & Backend
Knowledge
Media
