Content for humans,
wiki for agents
Make your knowledge discoverable by AI. Publish knowledge that people can read and that AI can understand, cite and recommend — from one source.
8 pages published
How it works
One source. Multiple representations.
You write in plain language. The page a person reads and the record an agent cites are generated from those same words, so they can never describe two different people.
Sections
For readers
Interfaces that explain themselves
Have you ever had to read a manual to use an app? The answer is almost always yes, but that says more about the interface than about the person using it.
Highlights that light up as the reader reaches them.
What AI sees
Marina Duarte
Product Designer
Expertise
- Design Systems
- Accessibility
Articles
- Interfaces that explain themselves
llms.txt — the profile as markdown
JSON API — the same record, structured
Person · ItemList — schema.org on the page
How AI reads you
Make your knowledge machine-readable.
Not a second copy for crawlers. The page, the schema, the markdown and the API are generated from what you wrote, so an answer engine can quote you without inventing a version of you.
01
Human page
the profile a person reads
02
Structured data
Person, Organization, ItemList on the page
03
llms.txt
the same profile, as markdown
04
JSON API
the same record, for programs
05
AI agents
understand, cite and recommend
A knowledge layer
People, organizations, and what they know.
Content for humans, wiki for agents. The wiki is not a dump of the page: it is who someone is, what a company ships, and how those facts connect.
Person
A named expert: what they know, what they have built, what they have written, and who else will vouch for it.
João Thaira ├── expertise ├── projects ├── articles ├── opinions └── references
Organization
A company as a graph: products, people, projects and the expertise that ties them together.
ThairaHub ├── company ├── products ├── people ├── projects ├── articles └── expertise
Knowledge
An article that points at a concept, a person, an organization and a product, so an agent can follow the links rather than guess.
Article ├── concept ├── person ├── organization └── product
Built with Thaira
ThairaHub is built with Thaira.
This is how ThairaHub structures its own knowledge so AI systems can understand its company, products, people and expertise — the same publishing model everybody else on this site uses.
A page that is yours
Your signature, your photograph, your handle. The layout is fixed so it always looks composed, and everything in it is yours to change.
Writing that shows
Publish articles in sections. Where showing beats telling, paste a Thaira Vision recording and it plays in place, beside the words it illustrates.
Machine-readable knowledge
Structured data, a plain-text profile and a JSON endpoint, generated from what you wrote. No second copy to keep in step — and nothing for an agent to invent.
Questions
What is Thaira?
Thaira is a platform for publishing human-readable content and machine-readable knowledge, designed to make people, organizations and their expertise discoverable by AI agents. Each person or organization gets a profile and writes articles on it, and every piece is published from one source as a page people read and as structured data, markdown and JSON that AI agents read.
- Is Thaira an SEO tool?
- Not in the usual sense. SEO competes for a place on a results page; Thaira makes what you know machine-readable — who you are, what you do and what you have written — so an AI agent can understand it, cite it and recommend you. The pages are ordinary, well-formed web pages too, so search engines index them like any other.
- Who is Thaira for?
- People and organizations whose expertise should show up in AI answers: founders, consultants, researchers, creators and companies. A person publishes as a Person and a company as an Organization, each with its own articles.
- What does an AI agent read?
- Five forms of the same source: the page, with schema.org structured data; an llms.txt with the profile as markdown; an llms-full.txt with every article in full; a JSON record of the profile; and each article as markdown at its own address plus ".md".
- How do I know AI is reading me?
- Thaira records when AI crawlers and assistants — OpenAI, Perplexity, Anthropic and others — fetch your pages, and checks each visit against the operator's published addresses where there is a list to check. Your panel shows which AIs read you, and each article shows how many have.
- Is Thaira free?
- Publishing is free, up to 4 new publications a month. Paid plans start at $39, €39 or R$197 a month and add articles ThairaHub writes in your voice, which you approve before they go out, and a regular report of which AIs read you: thaira.ai/pricing
Make your knowledge discoverable by AI.