An agriculture-technology startup with roots in Colorado’s cannabis industry is using artificial intelligence to help farmers monitor and save potentially diseased crops.

Founded by two European artificial intelligence engineers, deepgreen‘s application uses cameras to observe and identify harmful plant problems. The startup was accepted to the Boulder, Colo.-based cannabis-focused accelerator Canopy in 2017, and has since been targeting customers in Minneapolis and other strong agriculture markets.

Agriculture technology has grown exponentially in Minnesota over the past several years. Organizations like Midwest Pantry and Grow North have popped up to support local food and agriculture entrepreneurs, and the state’s largest startup competition. The Minnesota Cup added a food/agriculture/beverage division in 2014. Late last year, Techstars announced plans to launch Farm to Fork – an ag-tech accelerator that will be based in St. Paul. The accelerator is currently scouting for startups for its first class.

Since graduating from the accelerator, deepgreen has broadened its focus to include 80 percent of general agriculture, CEO Colin Ferrian told Minne Inno.

“Right now we’re monitoring everything from tomatoes to hops and barley,” Ferrian said.

Some of the most common diseases crops contract include powdery mildew, leaf rust and net blotch. Using affordable cameras, deepgreen monitors these and other diseases and alerts growers to any issues.

Disease identification and prevention are particularly important for plants grown in greenhouses or other indoor facilities. Cannabis is typically grown indoors, and a single plant could cost $500, Ferrian said. A room of these represents a small fortune, and one disease could wipe out the entire population in a matter of hours.

Vertically grown plants also represent a unique challenge indoors. If a grower has numerous crops several feet in the air, Ferrian explained, it’s not easy to keep an eye on all of them. If one of these high-growing plants contracts a disease, it could contaminate its neighbors before growers notice.

Deepgreen is geographically spread out, both in and outside the United States. Ferrian is based in Boulder, but regularly travels to Phoenix and Minneapolis. Deepgreen’s chief technology officer, Maxime Clauss and chief artificial intelligence officer, Max Unfried are originally from France and Germany respectively. Today, both are based in Vietnam.

Ferrian met Clauss and Unfriend through Canopy. At the time, Ferrian was advising on mergers and acquisitions for an equipment manufacturer, and identified an issue with the way some ag-tech applications are distributed.

“We found that there were only a handful of companies trying to solve this problem, and most of them had the same business model,” Ferrian said. “They’re building their own cameras and software and fighting for every single customers. We thought there must be a better way.”

Deepgreen’s “special sauce,” Ferrian said, is that its API plugs into existing software, and uses affordable cameras to observe the plants. The startup hopes to work with larger agriculture companies to add its technology into existing applications. Deepgreen has been testing its tech indoor growers for almost three months, according to Ferrian.

Ferrian said that Deepgreen is not currently raising any funding, but raised around $285,000 in an oversubscribed seed round. The startup plans to open another round of funding later this year, he added.

“There’s so much talent wasted on using A.I. engineers to get people to click ads online,” Ferrian said. “There’s an opportunity here to build something impactful for a lot of people.”