This document proposes a cloud-based food image recognition system that uses artificial intelligence and machine learning. The system aims to help users get information about foods by recognizing images taken on mobile devices. It was trained on over 6000 images across 13 common Indonesian food types using a convolutional neural network model. The system architecture involves mobile apps that send images to a cloud server for recognition by the AI model. Recognized foods and usage statistics would then be returned to provide nutritional information to users and analytics for future predictions. Key limitations include the number of food types that can be recognized and the need for a large training data set.