From coffee beans to milk, this Singapore-based startup uses AI to determine ingredient quality

The world has more mouths to feed than ever before.

As the global population grows by roughly 1% year-on-year, the demand for food production increases. But churning out more food also carries an environmental impact and potentially compromises its quality. 

One Singapore-based company has developed new technology which aims to address this challenge.

Using its patented digital fingerprint technology, ProfilePrint can rapidly analyze the identity and quality of ingredients, helping agribusinesses save money and time. 

“If you look at the raw ingredient trade today, you typically have the producer which are the farmers, you have the group of traders who buy to sell and then you have the end buyer, who buy to manufacture,” said Alan Lai, CEO and founder of ProfilePrint, speaking to CNBC’s “CNBC Tech: The Edge.”

The problem, he explained, is a system of multiple rounds of physical food samples exchanged between buyer and sellers, replicated across “every part of the supply chain.”

“So, the amount of sampling work … is significant,” he said.

“ProfilePrint digitalizes food ingredients, so that buyers and sellers no longer need to ship samples physically, allowing us to have a significant reduction in logistics costs, overheads as well as carbon footprint,” added Lai.

The technology is, on the surface, relatively simple. Companies can create their own AI models on ProfilePrint’s platform, which makes an assessment on whether the raw ingredient is a good fit for what they are looking for. Traders can also build AI models for specific clients to determine whether the product fits their preferences, based on criteria such as looks or flavor. 

At its core is a wave detector in the materials, combined with AI. Machine learning means the technology analyzes food samples at a molecular level and recognizes food quality within seconds.

Lai demonstrated the technology by using coffee beans.

“Maybe there’s one or even two beans that has accidentally fermented wrongly or we have an insect inside the bean that causes the entire bag to taste terrible, but if you cannot pick out that one or two then you effectively will not be able to sell this lot meaningfully,” he said.

 “So, the molecules are captured, converted into a digital fingerprint that contains this information.”

At this point, AI analyzes the fingerprint to generate a price or judgement, combining industry knowledge, within seconds.

As well predicting the quality and flavor profiles, the company’s ingredient quality platform can offer their client various options when it comes to balancing quality with price. 

With its patented software, PowerPrint’s technology has led to huge advancements in solving one of the most challenging issues in the coffee industry; spotting defective products that cannot be seen through the human eye.

The end goal is to streamline the whole supply chain and ultimately cut out human error.

ProfilePrint is not alone within the still embryonic pool of food quality and AI startups. U.S.-based startup Aromyx combines biotechnology, data science, and artificial intelligence to help businesses match flavors and scents for product development.

Brightseed, a startup based in California, uses its artificial intelligence platform, Forager, to identify compounds in plants and microbes and understand their health benefits. These are then used to create health solutions for food, beverage, and supplements.