The growing prevalence of lifestyle-related non-communicable diseases has created an urgent need for intelligent, personalised digital health tools. Conventional dietary applications provide only static calorie counters and population-average recommendations, failing to adapt to individual physiology, goals, or evolving health context. We present NutriAI Pro, a full-stack AI-augmented nutrition intelligence platform that integrates a React/Vite single-page frontend, an Express.js middleware server, Firebase cloud services, and the Google Gemini 1.5 Pro multimodal large language model. The system delivers four intelligent modules: an AI health coach conditioned on the user's biometric profile, a Food Vision pipeline that estimates caloric and macronutrient content from meal photographs, a generative weekly meal planner, and a predictive health analytics dashboard built on the Mifflin–St Jeor equation and a composite Health Optimisation Score. Functional evaluation demonstrates a 93% food-recognition accuracy, sub-second authentication, AI coaching latency of 1.8–3.2 seconds, and a 100% pass rate across the defined test suite. The work contributes a replicable engineering blueprint for AI-augmented consumer health platforms.
Katare et al. (Fri,) studied this question.