Why it exists
The platform needs multilingual delivery without forking the entire content model. Translation support is implemented as document-level variants linked back to a source record.
Use this when
- translating a schema into one or more languages
- correcting machine-generated translations
- reading translated content in a doc or website app
- modeling language-specific document views in the CMS
What it includes
- translation microservice endpoints
- TensorZero-backed translation workflow
- translation patching endpoint for manual correction
- language-aware document retrieval in the docs app
- translation-aware database columns and indexes
Core workflow
- Select a schema and target languages.
- Translate source documents through the translation manager.
- Store translated records linked to the source document.
- Patch translations manually when needed.
- Read translated content through language-aware schema fetches.
System surfaces
- translation manager API
- CMS language document views
- docs app language-aware content loading
