How Sojourn rebuilds a journey from a pile of photos

Open your camera roll after a long trip and you’ll find a thousand photos in a single undifferentiated scroll. Somewhere in there is a story — arrivals, detours, the quiet afternoons — but it’s been flattened into one long list. Sojourn’s job is to give that story its shape back.

Start with what’s already there

Every photo records a timestamp, and most outdoor ones record a coordinate. That’s enough. Sojourn never needs you to tag anything; it reads the metadata your camera wrote automatically and works from there.

Find home first

Before it can recognise a journey, Sojourn has to know what isn’t one. It looks for where your photos cluster most densely over time — your gravitational centre — and calls that home. Everything that happens far enough from home, for long enough, becomes a candidate trip.

Cut the timeline where the gaps are

People travel in bursts. Sojourn walks your photos in order and watches for the seams: a long jump in time, a sudden leap across the map. Those gaps are where one journey ends and the next begins.

A trip is just a stretch of your life that happened somewhere else.

Within each journey it does the same trick at a smaller scale — grouping photos into days, and days into places, using the same logic of distance and time.

Name it, then get out of the way

Each place is matched against an on-device geographic database to earn a real name — a city, a region, a country — with no lookup leaving the phone. A title and route fall out of that naturally: Kyoto & Osaka, five days, two cities.

The whole pass takes a heartbeat, and it runs every time you open the app, so a journey you took yesterday is already bound by the time you look. No upload, no spinner, no waiting on someone else’s server — just your own travels, handed back to you with their shape intact.