FROM DAILY FOOD RECORDS TO PERSONALIZED NUTRITION INTELLIGENCE — A DATA PIPELINE FOR DIETARY PATTERN, NUTRIENT AND WELLNESS ANALYTICS - (NUTRI TRACK)

Abstract

Diet is one of the most important factors affecting long-term health, yet most people have only a rough idea of what they actually eat over a week or a month. Conditions such as diabetes, obesity, anaemia, and high blood pressure are closely linked to eating patterns, and advice from doctors and dietitians is often difficult to follow without regular feedback. This paper presents Nutri Track, a data pipeline and analytics system that turns daily food records into personalised nutrition intelligence by estimating nutrient intake, identifying dietary patterns, and relating them to wellness indicators over time. Users record their meals through a mobile application by selecting dishes from a searchable list, typing free-text descriptions, scanning packaged food barcodes, or taking a photograph of the plate. The ingestion layer collects these entries along with optional data from fitness trackers, such as steps and sleep duration, and periodic measurements such as weight, blood glucose, and blood pressure. A food matching service maps each entry to a standard food composition database built from the Indian Food Composition Tables, supplemented with recipe breakdowns for common regional dishes, so that portion sizes can be converted into grams and then into energy, macronutrients, and micronutrients. The processed records are loaded into an analytical store organised by user, meal, and day. Daily and weekly aggregates are computed and compared with recommended dietary allowances adjusted for age, sex, activity level, and any health condition entered by the user. Clustering on meal-level nutrient profiles identifies each user's dietary pattern, such as a high-refined-carbohydrate pattern or a low-protein vegetarian pattern, and a gradient boosting model estimates the effect of dietary features on wellness indicators such as post-meal glucose and weekly weight change. A recommendation engine suggests practical substitutions drawn from dishes the user already eats. Nutri Track is intended for individuals who want to understand and improve their diet, and for dietitians who need reliable records of what their clients eat between consultations. It does not prescribe medical diets, and users with serious conditions are encouraged to follow professional advice.