SmartBites Culturally Aware Meal Planning for South Asian Foods

Abstract

Digital meal-planning systems depend on food records and dietary rules that reflect the meals their users eat. SmartBites is a web prototype designed around Pakistani and Indian foods, user-stated restrictions and a separate multilingual chatbot. This article describes its architecture and recommendation logic and examines the internal technical evaluation available for the prototype. The application combines a six-step user profile, the Mifflin–St Jeor resting-energy equation, activity and goal adjustments, tagged food exclusions and randomized selection among closely ranked eligible foods. We tabulated the specified test cases and local response times and checked numerical examples against the stated rules. Forty-one unit, functional, integration and security cases were marked as passed in the development record. Local point estimates were 1.2 seconds for plan generation and 2.8 seconds for chatbot response, without repeated-run distributions. For one female-coded profile, the printed expected resting-energy value of 1,239 kcal/day differs from the 1,289 kcal/day obtained with the stated equation; a meal-allocation example also mixes a fixed energy value with an incompatible percentage for most daily targets. SmartBites presents an implementable architecture for culturally aware meal selection.