Custom AI experiences

A build shaped around your organisation’s own knowledge and workflow, answering the way your own people do.

The moment

The question that arrives at eleven at night

A carer in a regional town types a question into your website after a long shift: what the new medication means for her mother’s morning routine. The answer sits in a folder on your shared drive. Your desk reopens at nine. By then she has rung the after-hours line twice.

Blank glass cards float in a ring around a bright core in a dark studio, with nobody in the frame. — Your policies, handbooks and recorded briefings become the material the build answers from.

Glowing message panels float in a dark studio, a ribbon of light winding away from them into the distance. — Your answer reaches her at eleven at night, drawn from the same material a daytime answer uses.

A glowing sphere hangs above a dark floor, rings of light spreading outward from beneath it. — One build reaches every site: in writing, by voice, or on a call.

What the build does

It answers from your material and hands over when it should

Retrieval draws on your approved knowledge, so the answer sounds like your organisation. Guardrails hold it inside what you agreed. Anything that needs a person is queued for one.

A practitioner moves knowledge into a tray. An AI roleplay scenario for custom AI experiences. — Your approved documents are the shelf the build reaches for, and nothing else.

Two colleagues review blank tokens for AI safety. An AI roleplay scenario for custom AI experiences. — Your safety team reviews what the build may and may not say before anyone outside sees it.

A tester uses the AI station's large control. An AI roleplay scenario for custom AI experiences. — Your visitors who cannot use a keyboard get a control they can.

Bring us the custom AI problem your team has.

Open the contact form

Before we build

What we agree with you before a line of code

Three things are settled before the first line of code is written.

Two people beside floating abstract shapes, one holding the answer given and one the intended flow. A first person AI roleplay scenario for applied AI in sensitive conversations, seen through the headset. — **A secure safety net.** A protected environment for sensitive conversations, so people get support on difficult topics.

Two designers compare abstract character shapes onscreen. An AI roleplay scenario for custom AI experiences. — **Data ownership and domain expertise.** Named vendors under their agreements, data handling you agree before we build, and your subject experts guiding what it can say.

A clinical scenario for practising conversation. — Your clinicians are coached through a session before the build reaches a patient conversation.

Three people arrange tiles for an AI workflow. An AI roleplay scenario for custom AI experiences. — Your team maps the workflow the build will sit inside, tile by tile, in the co-design workshop.

Our commitment, after it ships

What stays true once it is in your people’s hands

Tools that extend the reach of the people already doing the work, for one person or a whole organisation. Approved knowledge, human oversight, configurable boundaries and auditability stay in place.

Two people adjust a generic character. An AI roleplay scenario for custom AI experiences. — Your team adjusts the character’s look and voice until it fits the people it serves.

A builder presents an AI prototype for review. An AI roleplay scenario for custom AI experiences. — Your reviewers get a working prototype to try, not a slide deck about one.

Two specialists refine a translucent mask. An AI roleplay scenario for custom AI experiences. — Your specialists refine the model’s output until the edge cases are clean.

How a build comes together

From the discovery call to the day after launch

Discovery sets the goal. Co-design workshops shape it with your team. Prototyping builds it, testing readies it for your platforms, support carries it past launch. See our approach.

A tester raises an arm by a skeleton. An AI roleplay scenario for custom AI experiences. — Your testers put the build through its paces, and the skeleton shows what it saw.

A booking panel with a calendar beside the chat. A first person AI roleplay scenario for digital twins. — Your booking calendar sits beside the chat, so a conversation can end in an appointment.

A reviewer watches an AI conversation with the approval controls in view. — Your reviewer watches a conversation with the approval controls in view.

Where a person stays responsible

When the knowledge has to be yours and the call has to be human

A custom build needs curated, approved knowledge about your own policy and practice. Licensed professionals stay responsible for clinical and crisis care; the build extends the people already doing that work. Map where AI fits on AI Studio, answer a short set of questions and download the opportunity map, or start on the AI overview page.

A woman stands inside a glowing curved boundary with a translucent panel in front of her, a safety net drawn around a conversation. — Your people in a protected space see the companion, and the exit to a human, from inside the headset.

Inside a headset

The same build, in the room where the job happens

VR needs a headset. A custom AI build does not, but it can run inside one when the headset is already part of the job: a clinician in the staff room, a lounge host greeting a traveller, an AI presence beside the equipment. Same knowledge, same boundaries, same handover to a person.

A head wearing a mixed-reality headset, isolated against a soft field: one way a custom AI build can appear inside a headset session. — Your AI presence appears inside the headset session your staff already run.

A lounge scenario for welcoming a traveller. — Your lounge hosts rehearse the welcome with a traveller who arrives tired and late.

"AI plays a crucial role here as well, especially now, helping to script and gamify scenarios that provide a more realistic experience. It enables us to craft narratives that adapt in real-time, making the experience progressively more challenging and thus more effective."

Simon Lowe, CEO of People Tech Revolution, speaking at a podium. Simon Lowe
CEO, People Tech Revolution, on AI inside PTR’s scenarios. The Cusp of a Revolution, Global Best Practice Group, October 2024