Summary of how AI is transforming chocolate reformulation
- Cocoa shortages drive major brands to cut content and reduce costs
- Reformulation cycles are slow taking years from concept to product launch
- AKA Foods uses AI to shorten development timelines from years to weeks
- Platform unifies fragmented R&D data and integrates sensory insights for speed
- SMEs expected to adopt AI fastest due to resource constraints and data silos
Reducing cocoa content in confectionery has never been more popular. Industry giants like Nestlé, Mondelēz, and Pladis are all cutting back to offset the rising cost of a commodity hit by shortages in recent years.
But reformulation – and innovation cycles in general – are notoriously time- and resource-intensive. Developing a new product from concept to launch can take years.
Now, artificial intelligence start-up AKA Foods is slashing that timeline from years to weeks. Its new platform promises to help food companies create, optimise, and launch products faster and more cost-effectively than ever before.
Cutting cocoa content is a typical example
The software, coined AKA Studio, analyses a company’s own knowledge and ongoing R&D, as well as experimental and analytical measurements related to texture, aroma and taste, to help guide formulation and optimisation.
Of course, formulation remains a hands-on process, and the software won’t actually develop the product. But AKA Foods is still confident it can shorten timelines from years to weeks – whether that be to cut cocoa content, reduce sugar and fat in better-for-you products, develop brand-new innovations from scratch.
Cutting cocoa in chocolate formulation is a “very common” use case, explains AKA Foods CEO David Sack. “While AKA Studio doesn’t reformulate products automatically, it gives R&D teams the intelligence, context and alternatives they need to respond quickly when ingredients become scarce or expensive.”
That’s exactly what’s happened with cocoa. When a chocolate maker is facing this kind of shortage, innovating its way out of high prices is a potential solution. “The system helps teams search all past formulations, understand functional roles, and identify viable substitutes much faster and with more confidence.”
Addressing the biggest pain point for R&D professionals
Cocoa reformulation is far from the only demand manufacturers have. And across the board, for all innovation cycles, AKA Foods claims it can address the biggest pain point for manufacturers: fragmented R&D knowledge.
“Teams spend huge amounts of time searching for old formulations, sensory results and project files instead of building new products,” explains Sack. “When knowledge is siloed, time-to-market slows, reformulation becomes inefficient, and generic AI tools can’t help because they don’t understand food context.”
By bringing all that knowledge together, Sack says products can be developed and adapted much faster – and importantly, with fewer “blind spots”.
But actually, unifying R&D knowledge is just one problem for R&D professionals that AKA Foods says it can solve. They also say they’re looking for better integration of sensory insights into real formulation workflows, as well as wanting AI that is food-specific: understanding ingredients, process constraints, and formulation.
Are food manufacturers willing to hand over their IP to AI?
For AKA Studio to perform at its best, food businesses must grant access to their intellectual property and past R&D work. But in an industry famously protective of proprietary research, are players really willing to let an AI platform in?
That challenge is not new to AKA Foods. “It’s a concern we’ve heard before,” reveals Sack, assuring us that trust and security are central to the platform design. “Every customer operates in a fully private environment where their data is never shared and never used to train any external model.”
Clients can also choose an on-premise deployment if they so wish, whereby the entire system runs inside their own infrastructure with no external connection. “This gives businesses complete confidence that their IP remains fully contained and under their control.”
As to which kinds of businesses AKA Foods expects will seize its AI opportunity, it’s perhaps unsurprising that it’s predominantly small and medium-sized manufacturers. The bigger names have often already developed in-house systems. And SMEs tend to be more acutely affected and fragmented data and slow development cycles, says Sack.
AKA Foods is already partnering with manufacturers across the US, Europe, and Asia, though it remains tight-lipped about specifics. What is known, is that in the months ahead we can expect to see new product launches – or reformulations – powered by its platform making their way to shelves.
