Their name, in their voice.

The whole thing is a handover

1. You meet someone whose name you want to get right.
2. You hand them your phone and ask them to say it.
3. They say their own name. That recording is the record — yours is never played as their name.

You can add your own note beside it — how you heard it, where you met. It stays labelled as yours, and their voice always supersedes it.

Names are not all hard in the same way

A name with no marks in it can still be said wrong. The point was never that some names are difficult — it is that the person who owns one is the only one who can settle it.

What this app refuses to do

You cannot file someone who was not there. There is no way to type in a name you heard across a room, no way to look someone up, and no way to record a guess and drill yourself on it.

That is a smaller product, and it is the point. A system that accepts pronunciations only from the people they belong to can promise correctness. One that accepts them from anybody can only promise effort — and if the capture was wrong, practice makes you more confidently wrong every week.

Where it lives

There is no account, no sign-in and no server. Recordings stay on the phone that made them. Your backup is an export you hold — a file, encrypted with a passphrase you choose — not a copy of your contacts sitting with us.

Who tends to want this

• People whose own names get mangled, and who would not do it to someone else.
• Diaspora and immigrant communities, in both directions.
• Teachers and coaches at the start of a term, with a new roster.
• Clinicians, and anyone whose work is mostly first meetings.

Coming to iPhone and Android

It is being built now. There is nothing to sign up for and nothing to be added to — when it ships, it will be in the two stores.