It's a Friday night in the Waikato. The rugby's on, his wife's gone to bed, and Matthew Zonderop is staring at a laptop full of red error messages. Five weeks of mating spreadsheets — 400 cows' worth of decisions — just collapsed because of a single spelling mistake. It's 10:30 at night. He has to milk in a few hours. And in a moment of pure "what have I got to lose," he uploads the broken file to a chatbot he barely understands.

He didn't go to bed until two in the morning. By then, everything had changed.

This is the story of how a dairy farmer with no coding background, no startup money, and no plan accidentally built a tool that's now bending New Zealand's national genomic trendline faster than the breeding giants' own software — and what it means for every producer still drowning in data they can't make sense of.

TH ...