eWizard knowledge base

Appearance

August 10, 2026 ​

Improvements ​

  • eVa AI now recognizes claim-related and informational requests, improving response relevance.

  • You can now use the Refine text quick suggestion button in the eVa AI chat to instantly generate replacement-ready text for the selected text component on the content item layout. The generated text is based on documents from your knowledge base and considers the template metadata and content context. This streamlines creating text variants with minimal editing. Configuration required

  • In emails, email fragments, and email briefs, you can now specify how images captured with the Desktop view image option are displayed in the mobile version of the content item. Select whether to keep images at the desktop width or scale them proportionally to fit the width of the mobile screen.

  • In the Edit image text popup, use the new Fix background button to reconstruct the background in the areas occupied by text placeholders so that it matches the surrounding image. This reduces the effort required to update image texts, resulting in a clean, natural-looking image. Configuration required

  • Image text recognition now groups multiline text into a single placeholder when all lines have the same font size. This preserves the original text structure and streamlines editing by reducing the number of placeholders. Configuration required

Admin setup ​

  • To configure text refinement in the eVa AI chat, in the environment configuration file → aiAssistant add:

    • selected_text_generation value to the eva → supervisor → editSet array to activate the corresponding AI agent.

    • object labeled Refine text to the ui → hotActions array to make the Refine text quick suggestion button available in the eVa AI chat.

  • To show/hide the Fix background button in the Edit image text popup, turn on/off the Fix background with AI permission in the USER MANAGEMENT → Permission sets → SERVICES → Image text editor section.

  • To group multiline text into a single placeholder in the Edit image text popup, add the rowMerging parameter and set it to true in the environment configuration file → imageEditor → text-recognition.