FinNutriAgent (FNA): An Agentic AI for Nutrition Planning Considering Budget Constraints

This paper presents an agentic Artificial Intelligence (AI) model that is price responsive and combines household budget constraints and dietary optimization. Based on income, fixed financial commitments, health-related requirements, and dynamically revised food expenses, the system produces meal plans that are nutritionally balanced, cost-effective, and respond automatically to market shifts. The proposed design is an implemented modular multi-agent system, which consists of specialized agents, budget planning, nutritional evaluation, price surveillance, and health-based personalization. These agents align themselves with a common body of knowledge and use food substitution graphs to maintain nutritional adequacy and to reduce spending. The assessment based on a representative household in Saudi Arabia shows 13-18% savings in food expenses in comparison to the fixed weekly menus, nutrient adequacy greater than 95%, and stability in response to simulated price shocks of ±20 - ±30%. These findings indicate that the framework is valuable to match the affordability with nutritional sufficiency and offers a scalable solution to robust household diet planning in line with the Sustainable Development Goals associated with Zero Hunger and Good Health.