AV++® Technology
Personalized Media

Alice®

Different ages. Different interests. A story for each visitor.

PERSONALIZED MEDIA, NOT ONE STORY FOR EVERYONE

One exhibit. Different ages, interests and ways to explore.

A young child, an older child, a teenager and an adult can explore the same subject without receiving the same presentation. Alice® adapts the media, vocabulary, emphasis and depth to the visitor’s selected age group, interests and level of detail.

The content itself can change, not just its translation. A curious beginner might choose a clear introduction. Someone fascinated by the science, history or engineering can follow that interest further. Language, accessibility and pacing are part of the same personalized experience, grounded in the venue’s approved Body of Knowledge.

Age-appropriate explanations

Use stories and vocabulary suited to younger children, older children, teenagers or adults.

Interest-led content

Explore the science, history, people or engineering behind the same exhibit.

The visitor’s chosen depth

Offer a quick introduction, a fuller explanation or a detailed exploration. Age does not have to determine interest or depth.

Alice®

Alice® personalizes what a venue says, shows and explains for each visitor. Media is tailored to age, interests, background, preferred level of detail and pace, with language and accessibility integrated into the presentation. The venue remains in control of the knowledge and interpretation.

Alice® selects, assembles and presents media through public displays, personal devices, audio channels and interactive systems. A visitor can follow a short introduction, explore a particular interest or spend longer with the detail. The Alice® Body of Knowledge is the content foundation: the venue defines the approved facts, sources, interpretation and rules that shape the experience. Alice builds on patented and patent-pending personalized media technologies.

Knowledge foundation

The Body of Knowledge

Curators, educators and subject-matter experts define the Alice® Body of Knowledge. It gives the institution control of the material from which an experience is authored or assembled. Where conversational AI is included, its responses are grounded in that approved material and operate within the venue's rules. Generating a presentation does not give an AI authority to change the institution's facts or interpretive position. Unsupported questions need a defined response, such as acknowledging the limit or referring the visitor to a person.

Venue-owned content
The Body of Knowledge is controlled by the venue. Its approved sources, editorial permissions and use boundaries are part of the specification. This is Curated Flexibility: presentations can adapt to each visitor within those approved boundaries. The institution remains the authority for the story being told.
Curatorial authority
Approved material grounds the narrative, including when AI helps with its presentation. Content rules define permitted topics, supported sources and the response when the available knowledge does not support an answer.
Depth levels and repeat visits
Visitors can choose a quick overview, a more detailed story or a deeper study. With opt-in membership continuity, a return visit can build on earlier interests and experiences instead of starting again.
A story in the chosen language
The knowledge base supports multilingual content and synthesized narration. The project defines the approved languages, translations, voices and delivery paths for its visitors.
Accessibility-aware delivery
Alice® adapts pacing, vocabulary, and structure for visitors with different cognitive, sensory, and educational needs. Simplified language modes, extended pacing, and audio description variants are configurable per exhibit and per visitor preference.
Governed integration
Where WorldModel™ is included, content decisions connect with the venue's consent, accessibility, audience and operating rules. The content, permissions and delivery options are defined together.
Privacy as an engineering requirement

Intelligence that stays with the venue.

Mad demonstrates Alice® working with a local large language model, or LLM. Conversational intelligence can run on computers under the venue's control, close to the knowledge and visitor context it uses.

This is part of the transition toward an experience in which sensitive visitor information and important venue data stay within the venue, or within an explicitly authorized group of venues. The aim is to keep useful intelligence close to the experience while preserving control over the information that makes it personal.

The complete delivery path matters. Knowledge storage, model inference, speech, avatar rendering and connected services all need to respect the same defined boundary. Local Alice intelligence demonstrates a step along that path; each project establishes which services run locally and what, if anything, is permitted to leave.

CheshireCat® brings the same attention to local recognition and visitor choice. WorldModel™ provides the wider architecture for the permissions and rules that connected intelligence must follow.

Stack relationship

Alice® in the AV++® Architecture

Runs on QuickSilver®
Alice® runs on standard QuickSilver® compute nodes. No proprietary server hardware required. The same infrastructure that runs show control and exhibit operation also runs personalized content delivery.
Works with CheshireCat®
CheshireCat® can provide the visitor context that has been permitted, such as language, interests, access preferences and content already experienced. A temporary visit and an opted-in member record have different retention and recognition settings.
Extends via OnBoard and Avatars
Alice® OnBoard brings interpretation to supported personal devices. Alice® Avatars bring a guide or character to participating displays and personal touchpoints. Both use the knowledge and continuity the venue has authorized.
Guide, educate and entertain

One visit. A continuing story.

An Alice® Avatar can accompany a visitor through a venue, welcoming them, helping them find an exhibit, explaining what they are seeing and bringing a story to life. At participating displays, interactive stations or supported personal devices, the guide can continue with the selected age group, interests, depth and language, and the parts of the story already explored.

Continuity can last only for the visit. A visitor who chooses a membership or return-visit experience can instead permit a defined set of preferences and progress to be remembered. Recognition restores the context the visitor has chosen to keep. Sharing between venues is a separate choice with a defined scope.

CaterPillar™ can add nonvisual wayfinding, while Lory® provides the specified personal media and access formats. Each contributes to a visit that remains coherent as the person moves through the space.

Capabilities

What Alice® Does

Age- and interest-aware media

Adapts media, narrative emphasis, vocabulary and detail to the visitor’s selected age group, interests and preferences.

Multilingual delivery

Delivers the approved story in the visitor's chosen language using the content and voice options specified for the venue.

Accessibility Modes

Audio descriptions, simplified language, and adapted pacing for diverse visitor needs.

Continuous storylines

Connects experiences across participating spaces, retaining visit context or opt-in member progress according to the chosen settings.

WorldModel™ integration

Connects content choices with defined audience, consent, accessibility and operating rules where the governance architecture is included.

Works on QuickSilver®

Runs on standard QuickSilver® nodes - no proprietary hardware required.

Patented
Alice® builds on the patented methods behind personalized media delivery, with continuing development across content, local intelligence and connected visitor experiences. The public IP register distinguishes issued patents, registered marks and pending work. View the IP register.

Personal interpretation throughout the visit.

Alice connects venue-controlled knowledge with OnBoard on personal devices, Avatars in the experience and the visitor’s permitted preferences.

Explore OnBoard ↗

Alice® is available through project delivery, technology licensing and defined integration work. The project selects its QuickSilver computing, authored content and any CheshireCat®, Lory® or WorldModel™ interfaces. Mad’s local-LLM demonstration establishes a working capability; the complete venue deployment is specified and qualified separately.

Engage Early → See It at Wonderland Start a Conversation
TECHNICAL DETAIL

Keep the institution’s knowledge in charge.

Open the engineering detailClose this sectionLook for the red circuit symbol for engineering detail.

The venue identifies approved sources and interpretation, defines editorial roles and decides which transformations are permitted. The team then authors or configures the selected language, depth and delivery paths and reviews their behavior with the institution.

Illustrative boundary: a visitor asks a question that the approved material does not answer. The configured experience acknowledges that limit or offers a referral. The content workflow must establish the response and test it with the selected model and presentation path.

Current demonstrations, existing delivery tools and new integration work have different scopes. A project review identifies what will be used, what needs development and what acceptance evidence is required.

ENGINEERING DETAIL

Specify it: Alice® configuration.

Open the engineering detailClose this section
Alice® configuration brief
ParameterSpecification scope
Content authorityThe venue owns and curates the Body of Knowledge, editorial permissions and use boundaries. Curated Flexibility allows variation within those boundaries.
Local intelligenceA working demonstration uses a local large language model on venue-controlled computing.
PresentationSelect age groups, interests, knowledge levels, depth, pacing and media variants, alongside language, voice, accessibility and optional avatar or OnBoard interfaces.
Privacy boundarySpecify knowledge storage, model inference, speech, rendering and any connected services as one delivery path. Each service has an explicit local or permitted external role.
AcceptanceReview age-appropriateness, interest and depth choices, source accuracy, language quality, content boundaries, unavailable answers, access requirements and the complete delivery configuration.