Skip to content
FoodEdge
News

Kalsec and 2nd Nature explore AI-led sourcing from food side streams

The ingredient companies will assess potential uses of AgWastePortal. A useful commercial signal would be a candidate that survives application, recovery and supply checks; the announcement reports an exploratory collaboration.

For R&D teams
Fictional researchers in hair caps and gloves working with amber extracts, bran and citrus peel.
AI-generated illustration for FoodEdge. Fictional ingredient research.

Kalsec and 2nd Nature have announced an initial collaboration to explore naturally sourced ingredients. The 28 October company release says they will consider how 2nd Nature's AgWastePortal and Kalsec's market and technical expertise could identify opportunities with functional and commercial potential. Read the original company announcement.

According to 2nd Nature, its AI platform maps agricultural ingredients, including food-processing side streams, with predicted functionality. Kalsec supplies taste, colour and food-protection ingredients. The companies describe an evaluation of possible opportunities, with performance, availability, scalability and cost among the considerations.

FoodEdge view: from a predicted function to a usable ingredient

For an R&D team, our reading is that the useful connection is between molecular discovery and an ingredient buyer's application brief. A prediction can direct laboratory work. The commercially relevant question is whether a candidate performs in the intended food, can be recovered consistently from its source and can be supplied within the required specification.

Those checks can change which discovery deserves attention. A strong predicted effect could become less attractive if recovering the compound requires an expensive separation step or if its source varies between batches. This is a proposed evaluation framework, with no such failure or success reported for the collaboration.

Evidence of a candidate meeting a defined application requirement would move the story forward. A subsequent demonstration of consistent lots and a credible supply route would answer another question. Keeping those milestones distinct gives buyers a way to evaluate AI discovery without treating a promising prediction as a qualified ingredient.

Sources and reporting

  1. Read the original company announcement

This US collaboration is included for its broader relevance to ingredient-development methods. It provides no evidence of regulatory clearance, sales or demonstrated performance in Europe or APAC.

Desk report on 2nd Nature's 28 October announcement. No ingredient testing or supplier audit was performed by FoodEdge.

AI tools supported source research and drafting during the editorial process.

DiscussionOpenClose

More to explore

All publications