Eelj: a fair turn for a village's one harvester

The problem

One harvesting machine serves a whole village, and it can only be in one field at a time. Everyone waiting is losing grain — ripe rice falls into the mud, a storm flattens an over-ripe field — but nobody can see who is losing it fastest, so the order goes to whoever shouts loudest.

The plan I wrote

The app works out how quickly each field is losing its crop right now, from the weather it has actually had and from a photo of the crop itself. It builds the machine's order and route for the day so the least grain is lost across everybody, and re-plans in seconds when the machine breaks down or rain arrives.

Every farmer's phone can re-work the order on its own and confirm it is really what the agreed rule produces — so nobody has to take the queue on trust. And on day one, a single farmer already gets an answer: the best days to cut, and what each day of waiting costs.

The build

AI coding agents built it from my written specification. My work was the specification, the tests it had to pass, and the review.

The proof

On both test devices the signed release build ran with a real field: the phone's camera photographed the crop, the queue was signed, and the machine's start and finish were recorded. All six screens worked, and results survived a restart.

The day's plan came out identical on the computer, the phone and the tablet. 17 of 17 automated tests passed, with zero crashes.

The launch

Live on Google Play, with its privacy page and screenshots for phones and tablets.