The study introduces an improved approach for suggesting recipes using deep learning and machine learning techniques in response to the increasing need for individualized and health- conscious cooking solutions. The proposed technique blends transfer learning with ensemble learning in a fluid two-step process to enhance the accuracy and practicality of recipe suggestions. The first phase entails feeding specific elements into the system while the second phase involves feeding specific elements into a transfer learning model that accurately categorizes the said elements. Then, a random forest algorithm transitions the list of features that have been identified to follow a natural progression in group learning. As the last step, the transfer and group learning form the basis of a complex, solid recipe program. The ability of the system to provide personalized recipes based on identified ingredients is its main characteristic. Transfer and group learning combine to improve the accuracy and reliability of recommendations and ensure a wide variety of delicious foods. The process is comprehensive and provides the user with immediate information on the health benefits of the recipe upon receipt of personalized prescription recommendations. This crucial component enhances the user experience by encouraging a knowledgeable and health-conscious attitude to food choices. By combining cutting-edge machine learning with nutritional expertise, this research not only creates the first more accurate and customized recipe suggestion system but also establishes the groundwork for a fundamental change in the way individuals approach their culinary preferences. The marriage of technology with health consciousness is a significant development in fostering a healthy relationship between modern food and individual well-being.