Need visitors to get answers when nobody is there?
A character in augmented or virtual reality can stand beside your visitors, hear their questions and answer out loud. It knows where it is standing and what they asked a minute ago, so its answers fit the moment.
The idea Our message has to reach people when we are not in the room
Try it yourself
Ask the guide something.
You are in a gallery’s sea room, and you told the guide on the way in that you only have twenty minutes. Ask a question and compare its answer with a plain chatbot’s.
A scripted example. We wrote every reply here to show the difference; no live AI is answering.
-
You “I’ve only got twenty minutes.”
-
You “What am I looking at?”
-
You “And the one behind me?”
-
You “Have I got time for the café?”
Ask the guide
Character or chatbot
Closer to meeting someone than using software
Both answer questions. What changes is how people ask, what the answer knows and how it reaches them.
| A plain chatbot | A conversational character | |
|---|---|---|
| How you ask | You type into a box. | You speak, as you would to a person. |
| How it answers | Text on a screen. | Out loud, in a voice that suits who it is. |
| Where it is | In a window on a website. | A figure in the scene. You can walk round it and catch its eye. |
| What it knows | The words you have just typed. | Where it is standing, what is around it and what you said earlier. |
| What it feels like | Using software. | Meeting someone. |
Only need a presenter to read a script? Explore AI Video Avatars & Digital Humans
How it is built
Three layers make one character
When someone speaks, the voice layer hears them, the language model answers in character, and the voice and the body deliver the answer together. We tune each layer to your brief.
-
A large language model
Knowledge and personality
Gives the character what it knows and how it talks, so it answers as the figure you have designed together.
-
A voice layer
Ears and a voice
Hears what people say and answers out loud in a voice that suits the character, so nobody types.
-
Real-time animation in Unity
Body language
Drives gesture, lip-sync and reactions while the conversation happens, so the character moves as you speak instead of playing a fixed clip.
Our gamification work shapes how the encounter holds people’s attention. Explore Gamification
When it fits
A guide, a practice partner or a host
We build characters for heritage, entertainment, healthcare and training.
-
Heritage and galleries
Visitors question a guide or a historical figure about what is in front of them and hear the answers in character, in their own language.
-
Training role-play
A virtual patient or customer plays the other side of a difficult conversation, so your team can practise it somewhere safe. Our AI-driven training simulations work this way.
-
Hosts in location-based VR
In a free-roam attraction, a character greets each group, guides them through and points them to what else there is to try.
We do not hand over an AI component on its own.
We design the character, the world it lives in and how people meet it, together with you. It builds on our Immersive AI work and our long history in virtual reality.
How we work
From a character sketch to a first conversation
-
Decide who people meet
We work out with you who the character is, what it knows and what people should be able to ask it.
-
Build the character and its world
Together we shape its knowledge, personality and voice, then animate it in real time inside the scene it belongs to.
-
Test it with real questions
We put it in front of the people it is for, listen to what they actually ask and tune its answers with you before launch.
Questions
Good to know
-
How is a character different from a chatbot?
A chatbot exchanges text on a screen. A character is a figure in a 3D experience: people speak to it, walk round it and hear it answer out loud, and it knows where it is and what they have said.
-
Can the character speak more than one language?
Yes. The language model and the voice layer can listen and reply in several languages, which suits heritage sites and training for international teams.
-
Where can the characters run?
On augmented and virtual reality headsets and in location-based VR venues. The same approach carries over to WebXR experiences in a browser and to other immersive formats.
-
Do you build only the AI, or the experience around it?
The whole experience, together with you: the character, the world it lives in and how people meet it.
Your idea
Talk it through with us
Tell us what you are trying to explain, launch or change. We will work out with you whether this is the right medium for it.
Also in Immersive AI
Other ways to tell it
-
AI Video Avatars & Digital Humans
Your message keeps changing, your audiences speak different languages and nobody wants another day in the studio. An AI video avatar is a presenter made once who reads each new script, in each language you need.
-
AI-Driven Training Simulations
A rehearsal written once stops teaching after a few runs. An AI training simulation changes the situation every time, plays the other person and scores how your team handled it.
-
AI-Accelerated Synthetic Media
Booking the presenter again for every language is slow, so most films only ever reach one market. With AI-assisted production, one consented recording becomes a version for every language and, when it helps, for every viewer.
Your turn
What’s your idea?
Tell us the idea you are trying to explain, launch or change, and together we will find the right way to tell it.