Fermentelligence – Can AI-optimised fermentation of agri-food side streams deliver the sustainable protein promise?

Meeting rising global protein demand within planetary boundaries necessitates a shift towards more resilient, circular production models. Fermentation of agri-food side streams offers a promising land-light alternative with possible economic benefits but must overcome significant barriers such as feedstock variability, distributed supply chains, and safety. This article proposes Fermentelligence (AI-optimised fermentation) as the critical enabler to unlock this potential. The article presents how machine learning and digital technologies can manage system-wide variance by highlighting key recent work in this area; including optimising complex feedstock blends, real-time process control, and economic and sustainability-aware decision-making. Furthermore, it highlights the strategic importance of decentralised regional hubs and the retrofitting of legacy assets, such as breweries and wineries, to lower capital costs. Finally, the article outlines a de-risking roadmap emphasising open data ecosystems, food safety evidence, and pilot demonstrations to accelerate the transition from linear agriculture to a flexible, AI-optimised circular bioeconomy.