Try 1 month for 99¢
agrograph

Agrograph claims to use machine learning to predict yields for any kind of crop, anywhere in the world.

A two-year-old Madison startup has made a bid to create what its leaders describe as a historical “holy grail” of agriculture: Software that accurately that predicts yields “for any crop, anywhere.”

What’s more, according to its co-founders, it’s doing so in a way that is unique in its precision: Agrograph claims to be the only company in a competitive ag-tech market that gives predictions down to the individual field.

“We wanted to use the unit of measurement that farmers use,” explained co-founder James O’Brien. “If you’re farming corn in Iowa, or you’ve got a broccoli farm in the U.K., you base your yield on a field basis.”

O’Brien first met his co-founder, the University of Wisconsin-Madison computer scientist Mutlu Ozdogan, through the preschool their children attended. When they got to talking, O’Brien learned about Ozdogan’s history of “tinkering” with crop prediction tools, and his interest in commercializing them.

Ozdogan had an extensive background in using satellite-acquired data to map and model things like irrigation systems. O'Brien, for his part, was a Madison-area business consultant with a history of working with agricultural technologists. He said the two instantly realized the potential for a partnership.

“It was kind of that chocolate and peanut butter moment,” he said.

The business they developed with a third co-founder, software developer Pavel Pinkas, is not entirely new: Predictive crop modeling has long been  a hot sub-field of agricultural technology. Agrograph claims to stand apart not just for its precision, but for its clientele. Instead of marketing its technology to farmers, Agrograph makes its software for other “underserved” businesses in the industry like crop insurers and grain distributors.

According to O’Brien, farmers already have a wealth of technological resources thanks to the “precision farming” movement.

“Every farmer has a thumb drive with decades worth of farm data, riding around in the ashtray of their pickup truck,” he said.

Companies that work with those producers, however, lack the same degree of data-driven insight, said O'Brien. Predictions for crop yields are often based on reports from individual farms, with few means of independent verification.

“You’re making tons of decisions based on anecdotal data,” said O’Brien. “It’s kind of a black box.”

O'Brien said he suspects that because Agrograph does not work directly with farmers, the company lacks the "feel-good" veneer other ag tech ventures have. However, he thinks that the company is nevertheless doing something important.

“We’re right here in the heart of (the agricultural economy). And we’re serving industries that are based right here,” he said. “We’re just taking things one step higher up the chain, in an underserved market.

The two-year-old company just raised $500,000 in seed funding. According to O’Brien, the company’s focus is now on scaling and getting its product to customers. That said, while he hopes to maintain that focus, he also approaches entrepreneurship with some degree of flexibility. There’s always the potential for a pivot: “GM may want to know about corn yields when making cereal,” he said.

Subscribe to Breaking News

* I understand and agree that registration on or use of this site constitutes agreement to its user agreement and privacy policy.

Erik Lorenzsonn is the Capital Times' tech and culture reporter. He joined the team in 2016, after having served as an online editor for Wisconsin Public Radio and having written for publications like The Progressive Magazine and The Poughkeepsie Journal.