A guide to unlocking AI-fuelled food innovation [Interview]

Today’s products increasingly need to be not only cost effective and tasty, but also healthy and functional. The challenge for formulators has been how to retrofit ingredients that have been around for decades into new products that deliver what consumers are looking for. “While consumer demands have changed, ingredients have not,” says Nuritas CEO and founder Dr Nora Khaldi.

To address this challenge, Nuritas is leveraging advanced AI technology to accelerate the discovery of clinically proven, health-promoting ingredients. At this year’s Fi Europe, Khaldi will provide examples of AI’s transformative powers and explain why industry cannot afford to ignore this technology any longer.

During her session, entitled ‘Revolutionising R&D: Unlocking AI-fuelled food innovation’, Khaldi will show how AI can be used in product development, to drastically reduce the time and costs involved in discovering new ingredients that resonate with modern consumers. “AI is not a thing of tomorrow, it is a thing of today,” she says. “It is going to be integral to everything we do. The faster that companies integrate AI into their operations, the more likely they will maintain their market share.”

The need for new ingredients

Khaldi is a mathematician and computer scientist, with a PhD in molecular biology. She has won numerous awards, and in 2017 was named Woman of the Decade in Business and Leadership at a Women Economic Forum-EU event. After beginning her career in molecular evolution, comparative genomics, and bioinformatics, she turned to nutrition.

“When I started working in this sector, I was gobsmacked by the need for new ingredients,” she says. “There was a sense that while food is supposed to be healthy and nutritious, cost and taste have increasingly been prioritised as the sector has industrialised and scaled up. A consequence of diminishing the nutritional aspect has been a health epidemic.”

If ingredients are the problem, then why can’t we simply replace them with new, healthier ones? A key challenge, says Dr Khaldi, is the sheer time and cost it takes to discover new ingredients. As a result, reformulated products often contain ingredients that have been discovered 70 to 100 years ago and do not fit for what consumers are searching for today, but retrofitted to that demand. 

“Ingredient discovery is very expensive, and requires a lot of expertise,” she says. “Formulators need to balance cost effectiveness and taste, as well as health and nutrition. The reality is that traditional ingredient discovery techniques are not equipped to do all this.”

Nutritas’ journey and key AI insights

Nuritas was founded in 2014 to apply AI, deep learning and peptidomics to this challenge. The technology enables Dr Khaldi and her colleagues to look at the trillions of molecules, more specifically short proteins called peptides, that make up food with the aim of improving health and longevity. At Fi Europe, Khaldi will discuss Nutritas’ journey and share some key insights. 

“From the outset, our thinking behind this was how we could make ingredient discovery smarter and better,” she explains. “First, we had to create and build up our own experimental data. This took years, and cost millions. This was necessary to start intelligently educating our AI system.”

A key advantage of AI, says Khaldi, is that it is incredibly efficient. Traditional ingredient discovery by contrast is sequential; once a promising molecule has been discovered, it will eventually move on to clinical trials. Only then – and after years of development – will problems become apparent; perhaps the molecule breaks down too fast and their benefit is lost.  

“AI is different because you can establish the parameters you are looking for at the beginning i.e. you only want molecules that obey certain criteria,” she says. “This is why we have a human trial success rate of 80%, which is unheard of. But this would not have been possible without years of amassing proprietary biological and clinical data together.”

During her session, Khaldi will focus on ingredient case studies, including peptides. Nuritas has amassed the world’s largest natural peptide library accounting for trillions of peptides, so far the company has fully characterized over nine million food peptides that it discovered and associated to a health or a functional benefit.

“The peptides we discover carry various health benefits, three examples of this are improving our energy, or making our protein intake more efficient, so that we get more efficacy out of our daily protein consumption, or to promoting better sleep. But the technology is also capable of producing peptide ingredients that promote functional characters of a product like sweetness or improved texture, or shelf-life extension,” she says. “Without AI, it would have taken us millions of years to discover some of our ingredients.”

Perhaps Khaldi’s key message at Fi Europe will be that AI is not something that can be ignored. Consumers want healthier products with functional benefits, and the only way to discover new ingredients is by harnessing technology.

“I will also discuss ways of integrating AI,” says Khaldi. “The bottom line is that companies not on this journey today are likely going to lose market share and growth opportunities over the medium to long term.”