This study explores the integration of machine learning-based recommender systems to enhance artificial intelligence (AI) applications within engineering pedagogies. The research proposes a recommender system leveraging state-of-the-art machine learning algorithms, such as BERT and named entity recognition, to analyze individual learning behaviors from the users (both students and teachers) and adapt instructional resources accordingly. The system can be put into real practice in this stage. As the AI trend becomes increasingly critical in higher education, especially in engineering disciplines, innovative instructional technologies are needed to personalize and optimize student learning experiences. The recommender system provides tailored suggestions with personal learning needs to both students and teachers, facilitating problem-solving, resource selection, and reinforcement of key programming concepts. A self-constructed dataset of student interactions from programming courses in engineering was used to train and evaluate the model. Survey feedback from teachers and students further confirmed the system’s effectiveness in supporting programming education by streamlining learning paths and addressing common challenges. Comparative experiments across two universities demonstrated that the proposed system significantly improves autonomous learning, student engagement, and the development of personalized study plans. This research highlights the transformative potential of machine learning-based recommender systems in engineering pedagogies, offering scalable solutions for AI-driven personalized education in higher education institutions.