Draft — under review, not yet in force.
Legal
Subprocessors
Last updated June 11, 2026 · Version 1 · Questions? contact us.
This page lists the third-party providers ("subprocessors") that process personal data on behalf of ePress Norden AB to run [fix]net. Because [fix]net is local-first, most of what you make never leaves your machine — this list covers only the network services described in the Privacy Policy. Each provider is bound by a data-processing agreement. For B2B obligations and the full data-processing terms, see the Data-processing agreement (DPA).
The subprocessors
- Supabase — managed Postgres and authentication. Processes your account identity, device registry, subscription state, friend and sharing metadata, and the Store database (pack ownership, install records, marketplace analytics, author keys). Region: EU ⟨confirm⟩. DPA: in place.
- Fly.io — application hosting for the account/relay backend and the Store API. Processes data in transit while serving requests; the relay holds graph events in memory only. Regions: backend in Amsterdam (EU); Store currently in Singapore — production region to be confirmed (EU-preferred). DPA: in place.
- Cloudflare — content delivery, static site hosting (Pages), DNS, and R2 object storage for pack files. Processes request data and stores pack assets. Region: global edge; R2 bucket region ⟨confirm⟩. DPA: in place.
- Paddle — payments and subscription billing. Paddle is our Merchant of Record and acts as an independent controller for the payment itself, not only as our processor: it is the seller on your receipt and holds its own copy of transaction and invoice data under its own privacy terms. We receive only the customer/subscription identifiers and billing state needed to grant features. DPA / MoR terms: in place.
- Email provider — ⟨TBD⟩ transactional and newsletter email (support replies, account email, double-opt-in newsletter). No provider is selected yet; this row will be completed, with its region and DPA status, before any email is sent through it.
- AI inference providers *(conditional)* — only if managed inference is enabled. When you use your own AI keys (the default), prompts go directly from your machine to the provider you chose and do not pass through us. If managed inference is offered and you opt into it, the named model providers will be listed here, and prompt content will not be used to train their models.
How this list changes
This page is updated before a new subprocessor begins processing personal data, and changes are recorded in the legal changelog. Because we have no live users yet, no notice clock has started; once there are users, material additions follow the 30-day notice in our Fair-change Policy.
What never leaves your device
For completeness: the full local app, your local files and exports, and any plugins, MCP servers, or AI keys you run locally involve no subprocessor at all — that processing happens entirely on your own machine. See the Privacy Policy and the EULA.