Pair, Build, Publish: the three-verb model for digital signage

Most signage platforms have dozens of concepts you must learn in order: devices, players, assets, playlists, layouts, zones, channels, schedules, roles. ChannelOS has three verbs. Everything else is a detail you reach for only when you need it.

Here’s what each verb actually does.

1. Pair — the TV joins, no app required

Open play.channelos.tv in any TV browser. The page mints a screen and shows a pairing code and QR. Scan it from your phone and that screen is linked to your account.

Why it matters: the TV is only ever running a web page. There is no media player to buy, no app to sideload, no per-device license. Android TV, Google TV, webOS, Fire TV, a browser on a stick — if it renders a modern web page, it pairs.

Pairing is also reversible and honest: a screen knows when it’s been unpaired or taken over, and says so, instead of silently going stale.

2. Build — describe it, or design it

You have two on-ramps to a show, and they’re the same artifact underneath.

Describe it. Tell ChannelOS what the screen is for — a café menu, a clinic waiting room, a shop promo — and AI drafts every slide, on-brand, in seconds. Attach PDFs, images or a webpage URL and it reads them into slides. The output is a real, editable show, not a flat export.

Design it. When you want the last 10%, the editor is right there: Canva-style drag, resize, rotate and group, with undo, copy/paste and keyboard shortcuts that match across every surface. One-tap background removal, rich text, live widgets, and free Pexels photo and video are built in.

Creation by description, with the editor as your fallback — not your starting line.

3. Publish — push, or program

Two ways to get a show on glass:

• Push a show straight to a screen for an instant result.
• Program it onto a channel — a named broadcast feed with a real timeline and dayparting (weekday/weekend, by hour, set once a week).

The key idea: what’s on a screen is derived from the show and the schedule, never pushed slide-by-slide. So when you edit a show, every screen tuned to it updates in real time — no re-publish, no device-sync wait. Edit once; it airs everywhere.

The board ties it together

Above the three verbs sits one live board: every paired screen with a real rendered preview, pulsing live status, and one-tap controls — play, pause, blank, mute, next, hold. It runs over a realtime channel on Cloudflare Durable Objects, so the board and the screens never drift, and actions hit the glass in milliseconds.

That’s the model. Three verbs to learn, one board to run the wall. Everything powerful underneath stays invisible until the moment you want it.