Present, not in the way.

Detach the circle and it's a borderless, always-on-top bubble that follows you across desktops. Controls stay hidden until you reach for them. Present the whole time, in front of nothing.

Presence you read at a glance — who's here, who's talking — then straight back to what you were doing. No window to manage, no grid of faces pulling at your attention.

When you do need to point at something, your pointer steps onto the shared view, underlines it, and leaves a note — like Gustav, up there. Light touch, then out of the way again.

Video and audio are peer-to-peer by default. If a direct link fails, an encrypted TURN relay can forward the still-encrypted media without decrypting it. With everyone’s consent, self-transcription stays on-device; optional cloud AI receives only the selected meeting and canvas context.

Pull up the AI brainstorming canvas and it appears for everyone — a live, synced node graph you build together during the conversation. Throw ideas in, let the AI connect them. Screen-share with pointer annotations stays out of the way when you don't need it.

Møt hard-caps at five and won't become a webinar platform. It's the size of a room where everyone still gets a word in — and nobody's presenting.