AI Avatars

Avatars are the heart of Museum AI. Each avatar is a digital guide with its own personality, voice, speaking style, and knowledge scope.

Creating an Avatar

From your dashboard, navigate to Avatars and click Create Avatar. You'll configure:

• Name — The display name visitors see (e.g., "Professor Hartley" or "Curator Sarah")
• Persona — A detailed description of the avatar's personality and expertise
• Speaking Style — How the avatar communicates (formal, casual, storytelling, etc.)
• Era — The historical period the avatar represents
• Voice — Choose from dozens of natural-sounding AI voices powered by AI
• Default Greeting — The first message visitors see when starting a conversation
• Safety Level — Strict or standard content filtering
• Languages — Restrict which of the 10 supported languages this avatar responds in (see Guest Chat → Supported Languages for the full list)

Knowledge Scope

Each avatar can be scoped to specific exhibits or the entire organization's knowledge base. This means a "Civil War Historian" avatar only answers questions about Civil War exhibits, while a "General Guide" can discuss anything in your collection.

Avatar Styles

Museum AI supports multiple avatar presentation styles:

• Text Chat — Clean chat interface with the avatar's profile image
• Audio Reactive Portrait — An animated portrait that responds to voice
• Video Avatar — Full video avatars powered by HeyGen or Tavus (available on higher plans)

Quick Replies & Follow-ups

Enable Quick Replies to suggest conversation topics, and Follow-ups to keep visitors engaged with related questions after each response.

Best Practices

• Write detailed personas — the more context you give the AI, the better the conversations
• Use speaking styles that match your museum's brand and visitor expectations
• Test your avatar thoroughly before deploying to kiosks
• Create multiple avatars for different exhibits or audience types (kids, scholars, general visitors)