AIDA: Platform for Intelligent Nutrition

The platform allows users to create a diet tailored to their individual needs and goals. Users receive personalized nutrition recommendations, enabling them to choose food that best fits their specific requirements and preferences.

The Platform for Intelligent Diet consists of the following components:

• Food and Nutrition Knowledge Repository
• AI (LLM) Virtual Nutritionist
• AIDA Mobile Application
• Platform for Connecting External Systems with LLM Model (LLMconnect)
• Implementation of the Dietary Advisor Service in the Web Shop

Food and Nutrition Knowledge Repository

The Food and Nutrition Knowledge Repository provides users with comprehensive, relevant, and verified information on nutrition, food products, and recipes. The database of food products includes detailed information on ingredients, including their nutritional composition and a score based on it, environmental impact through CO2 emissions, information about specific product labels, seasonality, and methods for safe storage and use. The recipe database contains food preparation information and allows recipe customization according to users' specific needs, such as special diets, allergen avoidance, recipe adjustments based on ingredient availability, or preferences for a particular type of cuisine. Recipes are accompanied by nutritional value information, considering preparation methods. The nutrition database contains information needed for setting and achieving nutrition-related goals (nutrition recommendations), guidelines for sustainable eating habits, tips for reducing food waste, and up-to-date information on product recalls. Key elements of the database include a model for assessing food quality based on nutritional composition and a model for calculating the energy value of recipes.

AI Virtual Nutritionist

The Virtual Nutritionist is a digital tool powered by artificial intelligence that provides users with personalized dietary advice based on preferences, habits, dietary needs, and health recommendations. It is based on LLM methods of artificial intelligence that emulate the expertise of a real nutritionist and consultant. It possesses communication, professional, and analytical skills characteristic of the nutritionist profession while always being available to users at significantly lower costs. The system's primary function is enhanced by its educational skills, allowing users to monitor their daily calorie intake, access recipes that meet their requirements, check for ingredient alternatives currently unavailable, and more. The analysis of ingredients, composite products, and meals is enabled through the system's connection with the Food and Nutrition Knowledge Repository, allowing it to adapt every recommendation—whether in ingredient selection, quantity, or composition—to users' needs. The key principles guiding the system are providing answers and recommendations based on healthy eating principles, food safety, sustainability, and personal preferences, as well as offering scientifically backed responses rooted in proven nutrition principles.

AIDA Mobile Application

The AIDA mobile application delivers the services of the AIDA Virtual Nutritionist via mobile devices, allowing users easy access to food information and personalized dietary advice. By inputting personal data, health restrictions, and dietary preferences, users receive tailored recommendations that meet their needs. In addition to offering a personalized experience, the app plays a key role in collecting feedback on user behavior, questions asked, and satisfaction with the responses. This feedback is used for the continuous optimization and enhancement of the LLM model and other system components. The app's innovation lies in its user behavior tracking system, which not only records user activities but also actively asks additional questions related to the conversation topics through communication with the LLM model. This approach enables a deeper understanding of user needs and identifies key points for app optimization, delivering an even better user experience.

Platform for Connecting External Systems (LLMconnect)

LLM Connect facilitates easy integration of applications and websites with AI models like the Virtual Nutritionist. Once developed, the integration remains functional even in the case of model upgrades, changes to its communication interface (API), or a switch to a different model, eliminating the need for redevelopment. In the world of large language models (LLM), new models and enhanced versions frequently emerge, complicating the adaptation of existing systems. The traditional approach requires significant code changes with each model update. LLM Connect acts as a middleware between the application and the model, allowing users to quickly benefit from new models without additional investments or changes to the application. This approach optimizes the user experience by providing flexibility and staying aligned with technological trends. The product also enables adding new functionalities at minimal costs. In the case of the Virtual Nutritionist, users can easily implement the solution in their systems, ensuring rapid adaptation to future technological advancements with minimal costs.

Implementation of the Nutrition Advisor Service in the Web Shop

An AI Virtual Nutritionist is implemented in the web shop of organic healthy product manufacturers to offer the Dietary Advisor service. This provides customers with a personalized shopping experience, a new source of knowledge, a simple communication channel, and insights that help them make choices. This enhances customer engagement during shopping, ensures a better user experience, and increases the likelihood of purchases on the website. Implementing the Dietary Advisor service in the web shop of organic healthy product manufacturers aims to maximize the potential of the Virtual Nutritionist. This service is easily integrated into the existing web shop through the LLM Connect platform, which enables the connection of applications and AI models. End customers gain additional value through new sources of information and recommendations on healthy eating, methods of ecological farming, and the impact of the entire process of food production, processing, and consumption on health, the environment, and society. Additionally, customers are provided with suggestions for product usage, combinations, and recipes for specific meals. Customers receive a personalized shopping experience that helps them understand and choose products. The service also offers an easy communication channel with the nutritionist, increasing engagement and satisfaction. This approach improves the user experience and increases the likelihood of online purchases by offering a deeper and more informed experience than traditional website information.