Repo-native localization that never leaves the EU.
Sovereignty isn't a compliance checkbox we added later. Dialecto was built to hold as little of your data as possible, and to hold that little under EU jurisdiction — with a way out of even that, if you want your strings to never leave your own machines.
What we take, and where it goes
Your codebase stays in your repo. Dialecto receives a surgical slice — your translation files plus the usage snippets we need to be precise — and that slice is processed and stored in the EU. Nothing about your application beyond it ever reaches us.
Machine translation can stay local
Machine-translation suggestions can run on a local model, so your strings never leave the machine by default. No third-party AI vendor sees your content unless you decide one should — bringing a frontier model is your call, made per repo, never a default we quietly flipped.
EU-sovereign infrastructure
The app runs under EU jurisdiction with no customer data on US-controlled infrastructure. Its home is app.dialecto.eu, on an EU country-code domain, the same one this marketing site lives on. The site is a set of static pages with no cookies and no scripts, so no customer data is ever served from it.
Vendor independence
Everything rides on open protocols — S3-compatible storage, Postgres, SMTP, OAuth — so no single vendor, and no single US point of failure, sits underneath your localization workflow. That's the practical answer to CLOUD Act and FISA §702 exposure: not a paragraph in a privacy policy, but an architecture where the lever doesn't exist.
Self-hostable
Want it fully in your hands? Dialecto self-hosts on EU metal: Elixir, Postgres, object storage, and a CDN of your choosing. Same product, your keys, your jurisdiction.
Who this is for
EU and regulated teams with GDPR or data-residency obligations — and anyone who has read enough subprocessor lists to prefer a tool that keeps the list short.