Skip to content

This is an English translation. If it differs from the Dutch version, the Dutch version prevails.

What we keep about you, and why

This product measures the market around your products every night. That involves little personal data, but some, and all of it is listed below. Without template language.

Last updated 25 October 2026 · HG E-Commerce Online BV, Chamber of Commerce no. 84623969

Who is responsible

HG E-Commerce Online BV, Bagven Park 16, 4838 EH Breda, the Netherlands. Chamber of Commerce no. 84623969, VAT NL863285806B01. Questions about your data go to info@egantic.ai.

If you use Egantic as a seller, you are responsible for the data of your own buyers and we are its processor. What we may and may not do with it is in the data processing agreement.

Which personal data is in the product

  • You and your colleagues. Name, e-mail address, your role in your workspace, and on which days you opened the app. To let you log in, to manage permissions and to see whether the product is used.
  • Your buyers. Only if you use the customer questions module: the name the buyer gave, the order number and the text of the question, including whatever else is in it. To be able to draft a reply. Here we are the processor and you are the controller; that is what the data processing agreement is for.
  • Sellers you compete with. The company name bol shows with an offer. To be able to say who ranks on a search term. This is public information on bol and concerns companies, not people.

By far most of what we store is not personal data but market data: search positions, prices, numbers of sellers, review counts. That is about products.

Why we are allowed to

For your account and your market data: because it is necessary to perform the agreement with you. Without login details there is no product, and without measurements there is nothing to deliver.

For the customer questions: you pass them on to us yourself, and we process them solely on your instructions. We use them for nothing else.

For recording that you opened the app on a given day: legitimate interest. We want to know whether the product is used. It is one line per user per day, without path, time, device or referrer, and that is deliberately the minimum that answers the question.

Who gets to see it

We sell nothing on and do not use your data for advertising. These parties process data because the product does not run without them:

Supabase (Supabase Pte. Ltd.)

The database and the login details.

Location: Frankfurt, Germany (eu-central-1) · Everything that is in the product: your products, measurements, competitors, orders, margins, customer questions and user accounts.

Vercel Inc.

Running and serving the application.

Location: European region; the company is American · Data passes through server memory during a request. What remains are technical log lines with time, path and error message.

Inngest Inc.

Scheduling and running the nightly work.

Location: United States · Task data: which workspace, which product, which step. No customer questions and no personal data of buyers.

Anthropic PBC

The language models behind parts of the product, such as the weekly analysis and the draft replies to customer questions.

Location: United States · The text a request needs: product texts, figures, and for a draft reply the customer question itself. Anthropic does not use data from the API to train models.

Resend (Plus Five Five, Inc.)

Sending e-mail and receiving forwarded customer questions.

Location: United States · Recipient address and the content of the message.

bol.com (Bol.com B.V.)

The source of your sales and market data.

Location: the Netherlands · We read from bol with your own keys. Nothing changes on your bol account unless you confirm a change. Here bol is not our processor but your own sales channel.

For parties outside the European Union, transfers are based on the standard contractual clauses of the European Commission. Your database itself is in Frankfurt and does not leave Europe.

What happens with AI

A few parts of the product use language models from Anthropic, such as the weekly analysis and the draft replies to customer questions. Only what that one task needs is sent there. Anthropic does not use data from the API to train models.

The questions you ask in your own Claude or ChatGPT do not run through our account. The next section is about that.

When you connect your own AI

You can connect Egantic to your own Claude or ChatGPT. This only happens at your request, through a login at Egantic in which you choose the workspace. Such a connection reads your market data: search positions, visibility, prices and reviews of you and your competitors, search volumes, advertising pressure and the weekly analysis per product.

A connection can be read-only. If you give it write access, your AI makes small changes within fixed limits itself and prepares everything else for your approval. Every change is recorded with time and source in your log (Logboek).

What the assistant requests goes to the provider of that assistant (Anthropic for Claude, OpenAI for ChatGPT) under their own terms; we send nothing on our own initiative. Of each call we only keep which question was asked, when, and how long it took, not the answer.

You revoke a connection under Instellingen → Je AI (Settings → Your AI). A revoked connection stops working immediately.

How long we keep it

Measurements of positions, prices, offers and reviews

As long as you follow the product

The value of this product is in the series over time. If you remove a product, its measurements go with it.

History-import targets that are fully completed

60 days after completion

They have done their job once the history is in; keeping them only costs space.

Customer questions and their draft replies

As long as your workspace exists, or until you delete them

They contain data of your buyers. You can have them erased at any time; tell us and it happens within five working days.

The log of calls and changes

As long as your workspace exists

Which connection asked which question or changed something, and when. This is the record of what was done on your behalf; it should not quietly evaporate.

The usage overview (what each use of a language model inside the product cost)

As long as your workspace exists

It is the basis for your invoice.

Usage days (that you opened your work list on a given day)

As long as your workspace exists

One line per user per day, without path or time. Only to see whether the product is used.

Everything, on cancellation

Fully deleted within 30 days

You can still ask for an export during those thirty days. After that it is really gone, including from the derived tables.

What you can ask

Access, correction, deletion, restriction, portability of your data, and objection to processing based on legitimate interest. One e-mail to info@egantic.ai is enough; we respond within a month and usually sooner.

Deleting means really deleting. If you remove a product, the derived data in seven tables goes with it. If you cancel, everything is gone within thirty days; during those thirty days you can still request an export.

If you disagree with something and we cannot resolve it together, you can complain to the Dutch Data Protection Authority (Autoriteit Persoonsgegevens).

Cookies

Only what is needed to keep you logged in and to remember your light or dark preference. No tracking cookies, no advertising cookies, no third-party visitor analytics. That is why there is no consent banner: there is nothing to ask consent for.

Security

Data is separated per workspace, enforced in the database itself and not only in the application. Traffic goes over https. Your bol keys are stored encrypted and cannot be read back anywhere after you enter them. Access to the production environment lies with the builders and is logged.

If something goes wrong with your data despite all this, you will hear it from us, with what happened and what we are doing about it. Also if it is our fault.