Does scripture detection still work if the sermon is a recording, not live?

Yes. Kairos listens to whatever audio is actually playing in that room, live or recorded, and detects scripture from it the same way — there's no difference between a live sermon and a played-back one from the app's perspective.

Do all campuses need to show the exact same slide at the exact same moment?

No — and trying to force that over a network usually introduces more lag than it solves. Each site running its own local detection against its own audio feed naturally stays in sync with what that room is hearing, without needing a live network link between campuses.

Can each campus have a different operator with different settings?

Yes. Each site is a separate device activation under the same church account, with its own operator, its own manual overrides, and its own theme — while still pulling from the same shared song and scripture library.

← Blog

How to Set Up Kairos for a Multi-Site Church With Video Venues

September 11, 2026

TL;DR

A video-venue campus watching a recorded or simulcast sermon still benefits from live, local scripture detection and transcription — each site runs its own Kairos instance listening to its own room audio, so detection timing matches what that campus is actually hearing, not the original recording's timestamp.

Multi-site churches running video venues — a satellite campus watching a recorded or simulcast sermon from the main service — have a specific tech problem that single-site churches don't: the presentation software has to serve a room where the "live" moment already happened somewhere else.

Why Trying to Sync Slides Over the Network Usually Backfires

The instinct is to push slide changes from the main campus to every satellite in real time. In practice, this adds network dependency exactly where you don't want it — if a satellite's connection hiccups for even a few seconds, its display falls out of sync with a room full of people watching a screen, and there's no local fallback.

Kairos takes the opposite approach: each site runs its own instance, listening to its own room's audio — the same recorded or simulcast sermon feed that room is actually playing. Detection happens locally, against what that campus is actually hearing, which means it naturally stays in sync with that room without depending on a live connection to anywhere else.

Setting It Up Per Campus

Activate a device per site. Each campus's presentation machine is its own device activation under the church's license — Pro and Premium plans include multiple devices, with additional ones added individually.
Route the campus's actual playback audio into Kairos, the same way you'd route a live microphone feed — whatever that room's sound system is actually playing, clean, into the input Kairos listens on.
Let each site keep its own local settings — theme, output display configuration, and manual override behavior can differ per campus even though they're all pulling from the same shared song and scripture library.
Treat the video feed itself as a separate system. Kairos doesn't distribute the video/audio feed between campuses — that's handled by your existing streaming or recorded-playback setup (an NDI feed into OBS, a satellite broadcast system, or a simple recorded file). Kairos's job starts at "audio is now playing in this room," not before.

What This Gets You

Every campus gets scripture detection and live transcription timed to what its own room is actually hearing, not to a timestamp baked into the original recording. A campus running the Sunday 9am service's playback at 11am gets detection that fires correctly for an 11am room, without anyone re-cueing anything by hand.

It also means a satellite campus isn't a second-class experience — the same AI-driven detection and captioning the main campus gets live, the video venue gets from its own local feed, on its own schedule.