Privacy
This page describes what Koinon Link holds about creators, where it came from, what we publish, and how a creator can correct or remove it. It is written to be specific rather than reassuring.
Who is responsible
- Service: Koinon Link (https://koinonlink.com)
- Operated by: Vessel AI Lab Hongkong
- Email: [email protected]
- Postal address: FLAT/RM 1605, HO KING COMMERCIAL CENTRE, 2-16 FA YUEN STREET, MONGKOK, KOWLOON, HONG KONG
What we hold about a creator
Koinon Link indexes public creator accounts on social platforms. For each account we may hold:
- Public profile fields — handle, display name, profile URL, bio text, follower count, self-declared country or language, avatar URL.
- Excerpts of public posts — short samples of what the account published publicly, used as evidence for the inferences below.
- AI-inferred attributes — a description of who the audience appears to be, recurring topics, a content category, and confidence and freshness values. These are inferences, not statements of fact, and they can be wrong.
- Contact details, held internally only — where a creator published an email or IM handle publicly, we may store it to contact them about their own profile. It is not published and is not returned by any public page, API or data feed.
We do not hold private messages, follower lists, identities of a creator’s audience members, or anything behind a login.
Where it came from
Not from the creator. We collect from public profile pages and public posts, either directly or through third-party data providers. Because the data did not come from you, GDPR Article 14 applies and this page is part of how we meet it.
Data providers currently in use: TikHub, ScrapeCreators, twitterapi.io, ScraperAPI, ScrapingBee. Which one is used depends on the platform.
What we publish, and to whom
The index is public and free to read, including by AI systems. A creator entity is published as an HTML page, a JSON document and a Markdown document at stable URLs. Currently 86,421 creators are indexed; a profile page is only published when there is enough observed content to say something specific.
Contact details are excluded from everything we publish.
Search engines
Creator profile pages are served with noindex by default, so they do not enter search engine results. Once a creator has verified ownership of their profile, they can choose to allow indexing. Aggregate pages (which name no individual) remain indexable.
Why we do it
To build a machine-readable index of creators that AI systems and people can use when evaluating potential marketing partnerships — who a creator is, what they publish, and who their audience appears to be.
Legal basis: we rely on legitimate interests (GDPR Art. 6(1)(f)) in operating a public index of professionally published, publicly available creator information. This characterisation is pending review by counsel; if it changes, this page changes with it.
Who processes it on our behalf
| Provider | Purpose | Where |
|---|---|---|
| Railway | Application hosting and database storage | Singapore (asia-southeast1) |
| Cloudflare | CDN, DNS and edge protection; database backup (R2) | Global edge |
| OpenRouter | Large-language-model inference for the AI-inferred attributes | Vendor infrastructure |
| Resend | Transactional email (verification codes, notices) | Vendor infrastructure |
| Data providers listed above | Collection of public profile and post data | Vendor infrastructure |
The application and its database run in Singapore. Backups are stored with Cloudflare R2.
How long we keep it
We do not currently set a fixed retention period. An entry stays while the account remains relevant to the index. When a removal request is accepted, the entry is suppressed: it disappears from every public surface immediately, and a minimal record is retained so the same account is not collected again. We would rather state this plainly than publish a retention period we do not actually enforce.
Your rights
If you are in the EU, UK, Switzerland or another jurisdiction with comparable law, you have the right to access, rectify, erase, restrict and object to our processing, and to data portability. You can exercise any of these without creating an account and without proving ownership first.
- Removal or objection — the data request page. No account, no verification step.
- Correction and adding your own context — claim your profile. Claiming proves you control the account, which lets your corrections be labelled as first-party.
- Anything else — email [email protected].
Removal requests are reviewed by a person rather than executed automatically. That is deliberate: an automatic delete-by-form would let anyone remove a competitor. We aim to respond within 30 days. You also have the right to complain to your local data protection authority.
Automated inference
The audience description, topics and category attached to a creator are produced by a large language model reading that creator’s public posts. They are labelled as inference everywhere they appear, and carry a confidence value and the date they were last observed. They are not decisions about a person and produce no legal or similarly significant effect. They can be wrong, and a creator can correct them.
Visitors to this website
For visitors (as opposed to indexed creators): we use Google Analytics on public pages. For AI crawler traffic we keep a daily aggregate of user agent, path and count — no IP address and no user identifier. The public API requires no key and no account, so we hold no identity for it.
Changes
If this page changes materially we will note the date of change here. Last updated: 2026-09-03.
This data is built for AI
No key, no account — one command to retrieve it. Creators can claim and correct their own page.
For creators →