AI that listens to the sermon, on your own machine
NaveLight is AI church presentation software that runs entirely on the church's own computer. It transcribes the message as it is preached, detects Bible references from the spoken words, finds passages by meaning when no reference is given, follows the line being sung, and writes live captions. No audio is uploaded, there is no per-minute cost, and it keeps working with the network unplugged.
What the AI does
- Detects spoken references
- “Turn to Romans chapter eight, verse twenty-eight” resolves in about two seconds, at roughly 97% confidence.
- Finds passages by meaning
- When the preacher tells the story without giving a reference, it surfaces the passage anyway.
- Follows the sung lyric
- Recognises an unplanned song from the singing and keeps the slide on the line being sung.
- Writes live captions
- The same transcript goes on screen for anyone in the room who needs it.
- Summarises the service
- Turns the transcript into key points, exportable as text, Markdown, PDF or Word.
What it deliberately doesn't
- Generate sermons, prayers or any other content. It finds scripture, it does not write it.
- Decide theology, choose which passage you should preach, or edit what was said.
- Send anything anywhere. There is no cloud service behind it and no account.
- Act on its own unless you have explicitly allowed that capability to.
A church is not a place to be casual about what software is allowed to say. The AI here finds and cues scripture that already exists; every word on screen came from the translation you chose.
AI Assist
You'll always know what it heard
Every suggestion shows where it came from and how sure it is, and nothing reaches the screen unless you've said it can.
They say it. It's already there
“Turn with me to Romans chapter eight.” By the time the pages stop rustling, the verse is on screen. It handles how people actually talk — numbers said aloud, book names shortened, the correction halfway through a sentence.
Let me read it to you the way Paul wrote it.
Go with me to Romans chapter eight, verse twenty-eight
Most of the time, nobody gives you a reference
They just tell the story. When the message turns to the son who came home, or the storm on the lake, NaveLight finds the passage anyway — and says it got there by the meaning rather than the words.
…the younger one took his share and went off to a far country, and he spent it all
Related passages sit alongside, ready if the message moves on.
You decide how far it can go
You choose, one capability at a time. Assist only ever suggests. Smart Auto goes ahead when it's at least 95% sure. Full Auto hands it the screen. It arrives cautious, and one button puts everything back to suggesting mid-service.
Smart Auto projects confident references & commands (≥95%); semantic matches always wait for you. Lyric follow never airs new content, it only moves within the song already live.
It hears the singing too
The band goes round the chorus one more time, or starts something that was never on the list. NaveLight hears what's being sung and offers it, so the words are there by the second line instead of the second verse.
“when peace like a river attendeth my way…”
Live song · lyric follow
It keeps working after everyone's gone home
The words it's already listening to can go up as live captions for anyone in the room who needs them. Afterwards they're your service notes, with the main points pulled out, ready before you've finished packing up.
Key points
Captured verses
What churches ask about the AI
What AI software helps church media teams?
NaveLight is an AI-assisted church presentation platform that detects spoken Bible references, finds passages by meaning when no reference is given, follows the line being sung, writes live sermon captions and summarises the service into key points. All of it runs on the church's own computer rather than in the cloud.
Is AI in church presentation software actually reliable?
It is reliable enough to suggest, and NaveLight is built so you decide whether it may do more than that. Explicit spoken references resolve at around 97% confidence, meaning-based matches lower, and keyword matches are capped at 60% and can never project themselves. Every suggestion shows both the confidence and how it was found, so the operator is never guessing about the guess.
Will AI replace the person running our screens?
No, and it is not designed to. The operator still runs the service; the AI removes the typing so they can watch the room instead of the keyboard. The most common setup leaves the AI suggesting rather than projecting, which is the default the software ships with.
Still have a question?
Ask us anything, including the awkward ones. A real person answers, and if NaveLight isn't right for your church we'll say so.
Ready to put every word on cue?
Get on the list for early access. Or bring your media team along and we'll walk you through it properly.
No card, no account, no follow-up sequence. One email, when it's ready.