How AI for weeds and grains is fuelling this agritech startup’s global journey

Cgrain’s technology performs advanced visual inspections of grain, detecting everything from impurities and weeds to damaged or defective grains. It also measures physical attributes such as colour, size, and weight for each grain.

The firm’s primary customers are grain merchants and buyers at mills.

"The alternative that exists today is for one person to sit and examine each part of the grain manually. It’s time-consuming and monotonous work," says Moa Källgren, CEO of the startup since 2023.

From manual counting to automation

Cgrain’s journey began in 2013 when founder Jaan Luup developed an alternative to labour-intensive manual grain counting. Swedish agricultural giant Lantmännen saw the potential and contributed to developing the technology further.

"A typical example of its use is when Lantmännen buys malting barley from farmers – the barley is analysed with our technologies to determine the payment price," explains Källgren.

The company is part of Sting, one of Stockholm’s biggest startup incubators, and has secured early funding from Vinnova, Sweden’s innovation agency, and Almi, a state-owned fund supporting small and medium-sized businesses.

Fast forward to today, Cgrain has seen significant growth in recent years, with a 69% increase in turnover between 2022 and 2023.

Net sales for 2023 reached approximately €690,000. However, the company reported a loss of around €670,000 – a situation they now aim to address, including through a €370,000 capital injection from majority owner Lantmännen.

"In the first few years, up until 2015 when we sold our first system, the focus was heavily on development. Over the past three years, we’ve prioritised boosting sales and building an organisation capable of handling larger volumes. Our goal is to break even by 2025," says Källgren.

International expansion

Having established a foothold in Sweden and other European countries including France, Spain and Ireland, the company is now eyeing broader horizons. Turkey and South Africa are on the radar, but it’s the North American market that they’re most drawn to.

"If we make a breakthrough with leading players there, it could open new doors," Källgren notes.

Addressing new scepticism

Although Cgrain has utilised machine learning for grain analysis for years, recent debates about AI and data protection have brought unexpected challenges, says Källgren.

The company is increasingly having to clarify to authorities and international customers that its technology is limited to grain analysis.

"Our instruments are actually quite 'stupid' – they can only classify grains and are not self-learning," Källgren explains, adding:

"Just last week, we received questions from Canada about the limitations of our AI. It’s a new situation for us, having to emphasise what our technology cannot do rather than what it can. I understand the concerns about AI's societal impact, but our use case is a positive one – replacing subjective assessments with objective ones."

Despite the scepticism, Källgren sees significant potential to shape the industry and create a new standard for grain analysis.

"With our technology, we can establish new benchmarks for grain analysis across the entire value chain."