Documentation and education translation

Learning to translate learning materials

You lead the documents team for an online learning company. You're preparing the materials for a new course and need to translate the English documents into several other languages. That process has always been slow due to requiring manual copy-pasting and verification to ensure that only the desired elements are translated. That painful process repeats every release.

Translating documents is about structure as much as it is about the actual text. For that, you need more than sentence-level translation. You need a workflow that ensures headings, lists, placeholders, JSON fields, and content boundaries survive intact.

Match the format to the route

Lexel Translate offers native support for many documents types and routes depending on the content, including:

• Structured translation handles the case when the source already lives in a known format. content_mode accepts auto, plain, markdown, html, json, xml, yaml, and csv, and structured input is preprocessed before translation to ensure that only the parts that should be translated are translated (translate docs).
• Document jobs provide an all-in-one solution when you need to run translation on other file types. DOCX, PPTX, and PDF files are extracted to plain text, translated, and returned as translated .txt output (documents docs).
• Asynchronous jobs can handle your remaining translation needs, with a variety of options for callbacks and retention period to fit into your existing workflow (async docs).

Why teams choose this path

The value is operational:

• Content publishes faster, without the loop between the editor and a translation tool.
• Structured content comes back with headings, lists, and placeholders where they started.
• Larger document sets run as asynchronous jobs, and exports arrive as files instead of a series of copy-pasted translations.

Document jobs report their state through the lifecycle (pending_upload, upload_received, extracting, translating, assembling, completed, failed, canceled), so a large run is observable end-to-end rather than a fire-and-forget upload.

What changes

Course materials ship in every language the same day, with structure intact but without the hassle. Other documents render identically in each language and the documents team never holds up publishing again.