Cited, not claimed.
Citezen builds trust infrastructure for the age of AI answers. When an assistant answers instead of pointing at sources, the evidence has to live where the answer is formed. We build the systems that put it there.
What we build
Citezen is an umbrella for products that share one primitive — an evidence record captured at the source, sealed, and independently checkable — applied where being wrong is expensive.
- Audicta — audit-grade decision provenance for regulated AI. A separate solution at audicta.com.
- Citezen Destinations — AI-ready destination catalogs for long-tail tourism. First programme: Seychelles, with Tourism Seychelles, at seychelles.citezen.ai.
- Citezen Audit — the measurement instrument behind the destination work. In development; early access on request.
The name is citezen: to be cited is to be highlighted in the answer — a citizen of the AI web.
How we work
- 01
Measured, not guaranteed
We publish what we measured, with the date, the engines and the cohort. We do not promise rankings or outcomes, and we say "interim" when a number is interim.
- 02
Evidence at the source
A claim carries who stated it, when, and who checked it. Consensus between sources is not evidence; a statement from the party that knows, confirmed independently, is.
- 03
Sealed, then replayable
Records are written at the moment of the claim or decision, sealed by their hash, and can be replayed later without the system that produced them.
- 04
Tiers are visible
Trust has levels — T0 registry draft to T3 destination-verified — and the level is printed on every record, for people and for machines.
- 05
Nothing invented
If a registry does not know a price or an amenity, the field stays empty and the machine-readable notes say so. Absence is disclosed, never filled.
Destination authorities, DMOs and regulated-AI teams: [email protected]