Skip to content

June 16, 2026

Improvements

  • eVa AI now provides transparent source attribution for generated outlines. Each generated text is supported by visual indicators, allowing you to preview text derived from the source material in VVPM directly in the eVa AI chat. You can view detailed information—including source name, page number, and original text used for the outline generation—and navigate directly to the source in Veeva Vault PromoMats or the knowledge base. This enhancement improves content traceability, supports verification, and increases confidence in AI-generated email outlines.

  • The generation of assets.json has been enhanced to include VVPM asset usage metadata, providing visibility into where assets are used within content. This information can be used by publishing workflows and the MLR Acceleration engine to improve asset traceability and processing.

  • eVa AI now prioritizes references from selected source documents when generating interactive emails. This keeps the reference lists aligned with verified sources and improves consistency. Configuration required

  • With expanded asset type selection in the eVa AI chat, you can now search for and select specific claims and reusable text from Veeva Vault PromoMats (VVPM), as well as images from eWizard Library, VVPM, and NaviGate. Use selected assets to generate content that better matches your needs and is based on approved source materials. Configuration required

  • Enhanced eVa AI workflows across outline creation, content placement, and information handling:

    • A step-by-step guided flow drives email outline generation for interactive emails. eVa AI asks clarifying questions about brand, target country, audience, communication goal, or language if the initial prompt is insufficient.

    • Selected texts and claims are inserted into the most relevant email outline sections while preserving their exact wording.

    • Information retrieved from documents stored in a client-specific knowledge base or added with the Assets popup is now processed more efficiently. If content is missing or incomplete, eVa AI notifies you of the issue and suggests possible resolution options.

    • Relevant images are placed into email blocks based on image descriptions and block context when generating interactive emails. eVa AI informs you if not all selected images are used and stores your last image selection in chat history for later reuse.

    This improves content consistency, reduces manual effort, and supports more accurate and compliant interactive email generation. It also enhances decision-making by ensuring more reliable and context-aware content processing across emails. Configuration required

Admin setup

  • To activate the updated reference prioritization, add the "v2" value to the version field in the claimsReferences object in the environment configuration file → evamodules. If the version field is absent or set to a value other than "v2", the system falls back to the previous reference search behavior.

  • To control which asset types are available for selection in the eVa AI chat, add the assetsPopupActions array to the environment configuration file → evaui. The array supports the following values:

    • "media"—selection of images in the Assets popup.

    • "claims"—selection of claims and reusable text in the Assets popup.

    • "documents"—selection of document assets in the Assets popup.

  • To set up eVa AI workflows across outline creation, content placement, and information handling:

    • Configure the validate_email_request object in the environment configuration file → evamoduleshumanInTheLoop to manage the step-by-step guided generation of interactive email outlines.

    • Adjust the emailGuardrails object in the environment configuration file → evamodules to configure eVa AI for handling information retrieved from documents.

    • Add the relevanceSelection parameter with the true value to the selectedImages object in the environment configuration file → evamodulesemailLayout to activate relevant image selection based on image descriptions and block context for interactive email generation.