Data Processing Addendum

This addendum forms part of the Terms of Service and applies where UK or EU data protection law governs your use of MailDock.

1. The unusual part, first

Most software DPAs describe a vendor processing customer data. This one mostly describes the opposite. MailDock is licensed software that runs on infrastructure you control. The mail, the mailboxes, the users and the message content are processed entirely on your systems, and we have no access to any of it. In respect of that data:

• You are the controller. You decide why and how it is processed.
• We are not a processor of it. We do not receive it, store it, or have a route to it. There is no support tool, no telemetry and no back door through which we could.

Where we do process personal data — your account, your billing, your licence, your messages to us — we are the controller of that, as described in the privacy policy.

2. What we process, and on what basis

Data	Our role	Purpose
Account and billing data	Controller	Selling and invoicing a licence
Licence key, bound server identity, heartbeat times, mailbox and domain counts, product version	Controller	Operating the licence
Form submissions	Controller	Replying to you
Mail content, mailboxes, mail users	Neither controller nor processor	We have no access

3. AI features, specifically

This is the part a DPA in this market usually gets wrong, so it is set out plainly.

MailDock’s AI features run on inference you configure. When a user asks for a summary, the request goes from your server to the provider you chose, authenticated with your key. It does not pass through us and we have no record of its content.

Consequently:

• We are not a sub-processor for AI processing, and your AI provider is not our sub-processor. Their agreement with you governs it, and you should ensure it covers your use.
• You control what is sent. AI is per-user opt-in under a per-domain policy, attachments are excluded unless you allow them for a domain, and setting localOnly with a self-hosted model means nothing leaves your machine at all.
• The call log records metadata, not content — and only on your own server, which we cannot read. The admin copilot’s tool log keeps the arguments its tools ran with (a domain, a recipient searched for), on the same server.

If you need AI features that are contractually covered by us as a processor, that is a bundled inference tier and it does not exist yet (roadmap). Until it does, the honest answer is the one above.

4. Security

For the data we do process, we apply appropriate technical and organisational measures: encryption in transit, encryption at rest, least-privilege access, and no production access from personal devices. The measures protecting your mail are the ones you configure on your own server, and what the software provides is documented.

5. Sub-processors

Listed at sub-processors. We will update that page before adding one.

6. International transfers

Where personal data we control is transferred outside the UK or EEA, we rely on the UK IDTA or the EU Standard Contractual Clauses as applicable.

7. Assistance and breach

We will assist you, so far as we are able, with data subject requests and with any assessment you need to make. Because we hold no mail data, that assistance is mostly information rather than retrieval. We will notify you without undue delay of any breach affecting personal data we control that relates to you.

8. Signature

Pristine Technology LLC, a Texas limited liability company, at 815 Brazos St, Ste 500, Austin, TX 78701-2509, USA. If your procurement process needs a countersigned copy, ask through the contact form.