A product that classifies visitors should expect this question.
Here is the full answer. What we collect, what we infer, what we never store, where it lives, and who designed it this way.
- Collected: network level signals and page behaviour
- Inferred, not collected: industry and likely need
- Never stored: personal profiles without consent
- Where: EU, Frankfurt, under a standard DPA
- Built by: a founder with twenty years in information security architecture
What is collected
Network level signals of the visit: the network a visitor arrives from, the campaign token if one is present, and page behaviour such as depth and return visits. Datacenters, VPN exits, bots and consumer connections are filtered out rather than counted.
What is inferred rather than collected
Industry and likely need. These are classifications computed from the signals above, not facts gathered about a person. When confidence is low the system says so and does not guess. Language is not inferred, it is set once per client.
What is never stored
Personal profiles without consent. Company level aggregation runs on legitimate interest. Anything tied to a person exists only after that person actively agrees, and that consent is recorded server side.
Where the data lives
In the EU. A data processing agreement is included with every subscription as standard, and a departing customer is erased, with the documented, anonymized exceptions the agreement names.
Who built this
HyperW is founded by a security architect with twenty years in information security and cybersecurity architecture, including post quantum cryptography. The consent model, the filtering and the erasure path were designed by the founder, not retrofitted for compliance.
For the technical reader.
Identification
Visits are recognised at network level. Datacenters, VPN exits, bots and consumer connections are filtered out. A smaller honest number beats a large false one.
Consent
Company level first, on legitimate interest. Anything person level only after active consent, recorded server side.
Knowledge base is the source
We read your site, you approve the knowledge base. No invented claims.
Performance
Pages are pre-built and served from cache. Nobody waits on an AI response.
Integrations as they are
- LinkedIn: campaign links only. No data flows to LinkedIn.
- Microsoft Teams: alerts carry organisation-level data, never person names.
- CSV export of identified companies. No automatic sync.
- Campaign links resolve on our link domain or on your own. Either way no data flows to the channel.
Security
Hosted in the EU. Webhooks are encrypted, access is role based, and sensitive actions are audit logged. Controls are aligned with ISO 27001; HyperW is not certified. No certification claims we do not have.
Erasure
A departing customer is erased, with the documented, anonymized exceptions the agreement names.
Sending
HyperW never sends mail and never processes bounces or delivery feedback. Sending data stays with you.
AI transparency
Every funnel page carries the visible disclosure "Personalized by AI".
Recipient identity never reaches HyperW.
In personalized email campaigns the mapping from person to link stays with the customer. Our servers carry pseudonymous labels only.
Documents
Questions about any of this?
- Entity: Melya Holding B.V.
- Registration number: 51872587
- Address: Papendrecht, The Netherlands