Honest health information.
Finally.
When my daughter was diagnosed with vitiligo, I spent weeks searching for clear information about what people had actually tried. What I found was misinformation, predatory products, and no way to compare real experience against clinical evidence. WikiRemedy is the platform I needed and couldn't find.
- The problem
- The product
- Philosophy
- Design system
- Colour system
- Architecture
- AI collaboration
- Scale
- What's next
- Reflection
The gap between what the research says and what people actually tried.
When my daughter was diagnosed with vitiligo, I spent weeks searching for clear, honest information about what people had actually tried. What I found was scattered forums, strong opinions with no evidence, and no way to compare what worked for real people against what the research supported. This gap exists across almost every health condition.
What randomised controlled trials and systematic reviews say about a remedy's effectiveness. Reliable but often incomplete and hard to access.
What real people found helpful, for how long, and with what side effects. Valuable but unverified and scattered across forums.
WikiRemedy puts both in the same view.
Three screens. One coherent system.
The home screen surfaces trending conditions and the search. The condition page lists all remedies with category colour coding and dual scoring. The remedy page shows the full evidence profile, community ratings and the contribution flow.
Three principles. Every design decision.
Remedies are labelled with accurate evidence levels — including No Evidence for things widely promoted on social media. WikiRemedy tells users what the research actually shows, even when that's not what they want to hear.
Health is personal. WikiRemedy uses blockchain-derived anonymous identifiers so users can share experiences without any personal data being stored. No name, no email, no tracking.
Conventional medicine alongside Traditional Chinese Medicine, Ayurveda, naturopathy, herbalism and homeopathy — each tradition treated with equal respect while being honest about the evidence base.
Evidence and community. Two scores, one remedy.
Every remedy shows two signals. A Research score translates evidence levels to dot indicators — one dot for limited evidence, five for gold standard. A Community score shows star ratings from real users. The tension between the two scores is often the most useful information.
Category colour. Instant recognition.
Each remedy category carries a distinct colour that persists from the condition page through to the remedy detail — creating visual continuity and instant category recognition across 1,600+ remedies.
Canonical remedy architecture. One record, many conditions.
The most interesting technical decision was the canonical remedy architecture — a many-to-many database design where each remedy exists as a single canonical record shared across all conditions it applies to. Omega-3 for migraines and Omega-3 for dementia prevention share one canonical record but each has its own evidence level and community ratings for that specific condition.
One founder. A team's worth of output.
WikiRemedy was built entirely through AI-assisted development — no traditional development team. The entire codebase, database architecture, content generation and deployment pipeline was created through iterative collaboration with Claude. This represents a new model of product development: a founder with product vision and design thinking, working with AI as both technical partner and content engine.
People this would have required using traditional methods
Iterative AI collaboration throughout design, build and content
Built to grow.
The migration from one-to-many to many-to-many architecture was performed live without downtime — collapsing 1,352 remedy rows into 1,005 canonical records while preserving all existing community ratings.
The architecture is in place. The community fills it.
A chat grounded exclusively in WikiRemedy's canonical remedy data — not general AI training data
Powered by the canonical architecture — which conditions respond best to which remedies
Evidence review with specialist clinical advisers across each medical tradition
Founding community across condition-specific groups starting with vitiligo and migraine
It started as a personal problem.
WikiRemedy started as a personal problem — finding information for my daughter. It became a platform question: what does honest health information look like when community experience and clinical evidence are weighted equally? The answer is still being built. But the architecture, the design system and the product philosophy are in place.
Visit wikiremedy.org →