Avibord · Poultry farm performance tracking

Three chained failures. The producer does not measure, so they do not correct: daily mortality, feed actually served, and average weight at three weeks decide a cycle’s outcome, and are almost never written down. With no history, they are then invisible to the financial system: a technically excellent producer and a mediocre one present the same empty file. And because nobody records placements, the sector has no signal, so neither hatcheries nor feed mills know what is coming in three months.

The product hypothesis is that daily batch-level recording is the missing infrastructure, and that one gesture produces three effects: the producer’s technical decision, their credit history, and the sector’s aggregate signal. The model, 27 entities on Postgres, starts from the batch and its daily log (mortality, feed, weight samples) extended by protocol templates. It continues into listings, per-kilo bids and deals, with optional per-deal escrow, a ledger and payouts. Five applications: API, producer PWA designed for intermittent connectivity, server-rendered marketplace so listings are indexable, platform back office, marketing site. Producers are scoped by farm membership, buyers are not, so one person can both sell and buy.