AI Chef: Generative AI for Personalised Recipe Creation and Nutrition Planning

This paper presents an intelligent recipe generation system designed to support users in everyday meal planning. It recommends recipes based on the availability of ingredients and individual food preferences, addressing common challenges faced by users when time is limited or ingredient choices are restricted. To overcome these difficulties, artificial intelligence is employed through practical and user-friendly methods suitable for real-world cooking scenarios. The system monitors daily nutritional objectives, including caloric intake, protein, carbohydrates, and fat consumption, while also tracking water intake. Recipe recommendations dynamically adapt to user preferences, reducing monotony and simplifying daily cooking routines. In addition to creativity, the system emphasizes health-conscious meal suggestions. Users begin by entering ingredient details and meal type, which are processed by a backend server connected to an AI model responsible for generating appropriate recipes. The output is presented in a clear textual format to ensure accessibility for non-technical users. The Groq Cloud API is utilized to access the LLaMA 3.18B large language model developed by Meta, enabling rapid and efficient recipe generation. Generated recipes are displayed interactively, allowing users to review and analyze the dish. Extensive testing was conducted using various ingredient combinations, and Server-Sent Events (SSE) were implemented to enhance real-time interactivity. User feedback collected through a digital survey using google form indicates that the web application is effective and suitable for everyday use. The findings highlight the role of an online AI-based chef in supporting diverse user needs. The system also contributes to food waste reduction by encouraging the use of leftover ingredients instead of discarding them. Additionally, the application supports step-by-step recipe generation, enabling users to observe the formation process, thereby improving transparency and engagement. Integrated nutrition planning features track daily intake of calories and macronutrients, making the system more engaging and informative. Unlike traditional cooking applications that rely on static recipes without nutritional monitoring, the proposed system actively generates recipes while tracking dietary intake. By combining personalization, real-time interaction, and nutrition-aware recommendations, the AI Chef framework demonstrates how smart systems can effectively support daily cooking and meal planning in an accessible manner. The application is designed for ease of use and is suitable for a broad audience, including students, families, and working professionals, illustrating the applicability of artificial intelligence to routine household activities.