Skip to main content

Brand assistant

Interact with your brand guidelines conversationally.

What is the Brand assistant?

The Brand assistant is a conversational interface built into your brand portal. It allows users to ask questions, generate and review copy, and find assets, all based on the content of your brand guidelines.

Think of it as an extension of your brand team, available anytime to provide instant, contextual guidance.


What can you use it for?

You can ask the Brand assistant about anything within your brand portal, including:

  • Ask questions

    • “In which cases can I use a logo lockup?”

  • Review copy

    • “Review this text based on our tone of voice.”

  • Generate content

    • “Write copy for a product launch using our brand voice.”

  • Find assets

    • “Show me images with our logo and a bright color palette.”

  • Translate instantly

    • Ask in your native language and receive answers in the same language (100+ supported)


How does it work?

The Brand assistant uses a Retrieval Augmented Generation (RAG) approach. This means it generates answers based on your guideline content, not pre-trained knowledge.

  • Only text-based guideline content is used

  • The default language version is the source of truth

  • Responses include references to the original content


Who can use it?

  • Available to all logged-in users with access to the guideline

  • Can be enabled for both private and public portals.

  • In public portals, the Brand assistant is only visible to logged-in users. Visitors who are not logged in cannot access it.

  • Respects all targets and permissions


How do I enable or disable it?

Admins can enable the Brand assistant in Account settings, and then activate it for specific guidelines via the admin dashboard.

See full setup instructions in this help article.



Customize the brand assistant

Editors and Owners can customize how the brand assistant appears within individual guidelines by updating its name, welcome message, and disclaimer.

These settings allow organizations to align the assistant with their brand, legal requirements, and internal AI governance policies.

To customize the brand assistant:

  1. Open your guideline

  2. Open Guideline settings

  3. Navigate to the brand assistant tab

  4. Update the assistant name, welcome message, or disclaimer

  5. Save your changes

Assistant name

Customize the name displayed to users when interacting with the brand assistant.

If left empty, Frontify's default name is used.

Welcome message

Add a custom welcome message to introduce the assistant and guide users on how to use it.

Welcome messages support rich text formatting, including bold, italic, underline, and links.

If left empty, Frontify's default welcome message is displayed.

Disclaimer

Add a custom disclaimer that reflects your organization's legal, compliance, or AI governance requirements. Disclaimers support rich text formatting, including bold, italic, underline, and links.

For example, you may wish to:

  • Explain that responses are AI-generated

  • Direct users to official source material

  • Link to internal AI policies

  • Clarify that responses should be reviewed before use

If left empty, Frontify's default disclaimer is displayed.

Important note

If any field is left empty, Frontify automatically displays the default content. This ensures the brand assistant always provides a complete experience without requiring configuration.


How does it work with translations?

The Brand assistant supports 100+ languages and will respond in the language used in the prompt.

However: It always references the default language version of your guidelines. Content that exists only in translated versions will not be used


Can it find assets from my DAM?

Yes. The Brand assistant can retrieve assets based on their description fields.


What should I keep in mind?

  • Answers depend on your guideline content: The Assistant can only use what is written in your guidelines.

  • AI output is not always perfect: Always verify responses using the provided references.

  • Text-based only: Images, PDFs, and videos are not currently used.


Data and privacy

  • Your data is not used to train AI models

  • Guideline content is accessed securely to generate responses

  • Interaction data may be processed and temporarily retained by Azure for up to 90 days for monitoring and quality purposes.

  • No additional cookies are introduced

  • Personally identifiable information (PII) is only processed if entered by the user


Is the Brand assistant still in Beta?

No. The Brand assistant is now generally available and enabled by default for customers. We continue to improve quality and expand capabilities over time.


Feedback and support

Use 👍 / 👎 in the product or contact your Customer Success Manager.


Where can I learn more?

Did this answer your question?