Immersive stage configurator

• One room, no shared picture

  Teams talked about "the tablet by the window" with no common view, so placement mistakes surfaced on test day.

• Links typed by hand

  Every device needed a link typed against the laptop's address, and one typo or address change broke the session.

• No rehearsal without the room

  Seeing the flow across every screen took a full dry run in the room, so problems surfaced late.

• Sensitive guides, no uploads

  Interview guides and lab details can be sensitive, so nothing could leave the browser unless someone downloaded it.

1. Where does each device stand, and what will a participant see from there?
2. How does every device open the right screen without anyone typing a link?
3. Can a team check the whole flow, across every screen, before test day?
4. How do the scenarios in an interview guide reach the prototype without retyping?

A real lab has instruments, sample flow and people moving between screens. The immersive stage recreates it digitally in a controlled room: instruments projected on the wide wall, tablets held up as their screens, an overhead screen for status. That lets a team test a prototype in context.

Real lab. Photo: Tima Miroshnichenko, Pexels. The stage recreating it is rendered.

The floor plan runs in five steps. First, pick the touch points: moderator tablet, tablets, phone, desktop, wide wall and overhead screen. Then click a spot in a 3D model of the real room, and the device lands there. An eye icon shows the view from each kind of spot.

The plan tilts into 3D; a click places a device and a drag moves it.

Step three connects the prototype: its repository, where it runs and the route each touch point opens. This is where our agentic prototyping workflow plugs in: a prototype in a repository reaches every screen on the stage. For now it is a concept step, with no sign-in and no uploads.

Each touch point opens its own route; an empty route keeps the device's lab screen.

Every touch point gets its own link and QR code, so each device scans once instead of someone typing a link. Every screen can be previewed live, and on test day a host dashboard shows all the codes on one screen.

On the host dashboard, each card reads "not connected yet" until its device scans in.

Step five plays a scenario through the room, and the camera moves to each touch point as its screen changes. The team checks the whole flow across every device before the session begins, without a full dry run in the room.

Click any device to fly to it and see its screen at that step.

Personas and scenarios sit on the second tab as simple cards. Drop in the moderator's interview guide as Word, Markdown or text, and the configurator proposes scenarios mapped to the touch points for review. Then one Markdown file holds the whole lab: devices, placements, routes, personas and scenarios. Until someone downloads it, everything stays in the browser.

The Scenarios tab: personas, scenarios with options A and B, and the guide import.

• Rehearse before the room. Playing a scenario on screen finds the problems that used to wait for a full dry run.
• Make the set-up a file. One Markdown file holds the whole lab, so a session can be shared, reused and restored.
• Keep sensitive guides local. The interview guide is read in the browser, and nothing leaves it unless someone downloads the configuration.