Polyglot
Write once in English. Publish everywhere, in every language — fact-checked before you hit the button.
Most blogging platforms treat translation as an afterthought and fact-checking as someone else's problem. Polyglot was built on the opposite premise: a post isn't finished until it can travel, and it shouldn't travel until it's true.
An author writes in English. Polyglot renders the post into any number of target languages and locales — not just word-for-word translation, but locale-aware phrasing, so a reader in São Paulo and a reader in Lisbon each get prose that feels native.
The community keeps translations honest
Machine translation is good; native speakers are better. Any signed-in reader can suggest an improvement to a translated passage. Suggestions queue up for the author, who can accept them with one click. The same mechanism handles freshness: if a post's content goes stale — a statistic superseded, a product renamed — the community can flag and propose the update, keeping every language edition current instead of only the English original.
Fact-checking built into the publish button
Before publication, Polyglot searches online for sources relevant to the post and attaches references to the claims it can verify. When a claim fails verification — say the draft asserts the sun rises in the west — that section is flagged and shown to the author before publishing. The author stays in control: Polyglot never silently edits, it annotates. The result is a post that ships with its receipts attached.
How the backend works
A detailed architecture write-up is coming here: the translation pipeline, the suggestion/review queue, the retrieval system behind fact-checking, and how references are ranked and attached. Watch this space.
Keep reading
The full write-up — community translation, fact-checking, and the backend architecture — is free for members.