A relative starts a story. You press record. Your phone listens — and so does the archive, for the first time.
Speech becomes text right on your device, with speaker labels. Nothing is sent anywhere. Watch the transcript appear, word by word.
Where does Nana talk about the kitchen? What did Papa say about the war? Answers cite the recordings — and the exact moment inside them.
Your grandmother’s voice is the most private thing you’ll ever hold. So we built Family Museum the only way it could be built: everything that can happen on your phone, does. Everything that must be stored, is stored in your own iCloud — locked inside Apple’s end-to-end encryption, where even Apple cannot read it.
Speech-to-text, speaker labels, entity extraction — all executed by the Neural Engine in your iPhone. The audio never crosses the internet.
Recordings sync through iCloud with Apple's end-to-end encryption. The keys are stored in your Secure Enclave, tied to your Face ID.
We learn who sounds like whom — but the voice embeddings live on your device, not in a central database. They never leave.
Record at your grandfather's farmhouse where the signal's bad. Everything still works. Things sync when you're ready — or never, if you prefer.
Family Museum is free for anyone who just wants to listen. If you’re the one doing the recording, there’s a small subscription that keeps us independent — and keeps your family’s voices out of someone else’s training data.