Auto-translate your survey

Use the AI machine translations provided by AWS Bedrock to automatically translate your survey text. By auto-translating survey text as the first step and then verifying the translations afterward, you can increase authoring efficiency.

This feature is available for all supported languages.

Note:
  • This feature is available for communities in the NA2, EU1, and EU2 pods.
  • If you are in the NA1, AP2, or AP3 pods, you must agree to US data processing. Please contact your Customer Success Manager.
  1. In the Translations workspace, in the top right corner of the toolbar, click Auto-translate.
  2. If this is your first time using an AI feature, review the information in the Do you want to use AI? dialog.
    You must consent to Alida's terms and conditions, and enable AI and Machine Learning, before you proceed. For more information, see Enable AI and Machine Learning.

Once you click Auto-translate, your survey text is submitted for translation. The application detects whether there is new or updated survey text in the source locale, and whether there are existing translations. Translations are provided for new or updated survey text only. If you are updating survey text, you must first remove the existing translation strings so that the translation fields are empty; otherwise, the translation submission will not succeed.

If the text hasn't changed, the request is not sent.

If the application detects that new translations are needed, a notification message appears across the top of the page. Translations generally take a few minutes, but that duration depends on the server load at the time of submission. Requests are pulled every 30 seconds. Requests for multiple languages are submitted and returned at the same time.

When the translations are returned, the notification message disappears.

  1. Review the translated survey text.

    Machine translations are a good first step, but they don't replace the need for a fluent speaker to review the text. As you read the translations, pay special attention to context, vocabulary, tone, formatting, and any HTML tagging that might be wrong.