Karevo

Manual sorting of potatoes and onions is slow, inconsistent, costly, and physically demanding. Moreover, it is becoming harder and harder to find staff for manual sorting. When we began our journey, the price range of the existing automated solutions available on the market started at ~150,000 EUR, placing them out of reach for small and mid-size operations. Karevo delivers the first affordable, AI-powered optical sorting system designed specifically for family farms. Running entirely on edge AI, the system detects seven defect types plus foreign objects with ~95% accuracy, compared to ~80% for manual sorting.

• Building trust in a conservative industry. Agricultural technology adoption relies heavily on peer recommendations and proven reliability. As a young startup, convincing farmers to trust AI-powered automation required extensive on-farm demonstrations and partnerships with established names.
• Developing the machine. Developing a robust sorting machine that combines industrial-grade mechanical engineering with cutting-edge computer vision is not a trivial task. It takes a lot of time and even experts in the field could not help us anymore as we were entering unchartered territory.
• Hitting an accessible price point. Family farms on 5–50 hectares operate on tight margins. Designing a machine powerful enough to compete with 150,000+ EUR industrial systems while keeping the price at 75,000 EUR required ruthless cost discipline in component selection, operations and manufacturing.

• Provided us with a place to work, the ability to use the machines in the maker space, access to people with practical experience, a fridge with beer
• Offered visibility through TUM Venture Lab events and the broader TUM entrepreneurship network.
• Supported us in carrying out project studies together with students from TUM

Build something that solves a real, tangible problem. Spend as much time with your customers in the field as you do behind your laptop.