<p>This research presents an integrated framework for energy and sustainability optimization in food production systems, which is based on a combination of digital twin simulation, multi-objective fuzzy optimization, and an Agentic AI layer for contextual selection of operational decisions based on Pareto solutions. The framework evaluates the trade-off between energy consumption, operating cost, carbon emissions, and product quality under operational uncertainty. The empirical analysis of the research is based on data from an industrial food production line and is complemented by digital twin-based simulations to cover a wider range of operating conditions and decision-making scenarios. The results show that the proposed approach reduces energy consumption by about 6–8% and carbon emissions by about 5%, while maintaining product quality stability. These findings indicate that the proposed framework can support more reliable and implementable energy and sustainability decisions in complex food production environments.</p> Graphical abstract <p></p>