Nutrition is a modifiable lifestyle factor that has a fundamental role in human development and health. Recently, there has been growing interest in food recognition and nutritional analysis, driven by the strengths of machine vision based models to estimate portion weight, volume, and nutrition of food dishes. However, research has shown that relying solely on image recognition techniques may not provide accurate weight and nutritional information. The latest AI-based object detection algorithms have also enhanced the accuracy of food recognition and nutritional estimation. This study utilizes an AIoT system as Flavoria Flex to integrate popular AI based algorithms for food recognition, weight estimation, and nutritional analysis to compare their performance against ground truth data collected from the Flavoria restaurant’s lunch line. The AIoT platform helps in collecting and validating this data by combining AIpowered food recognition with scaled weight and menu-based information.