The Polyglot Engine provides infrastructure to load open access texts from LibreTexts libraries, pre-process and translate them, then save them to a designated path on another LibreTexts library.
As of the alpha release, it is designed to run on Amazon Web Services cloud infrastructure and use the Amazon Translate neural machine translation service.
The Polyglot Engine consists of one startup module and two main "cylinders":
- Ignition validates requests and adds them to the Engine queue.
- StartTranslation gathers content from the specified library, pre-processes it, then initiates the machine translation.
- ProcessTranslated retrieves the translated content, reassembles the original text's structure, then saves it to the designated target library.
For more information about the Polyglot Engine and how we use it, reach out to the LibreTexts team at [email protected].