The virtualization of university courses through artificial intelligence (AI) and big data offers a transformative pathway for enhancing higher education. This study develops intelligent professor avatars capable of delivering lectures and engaging in interactive dialogue with students by embedding processed course materials into conversational models. A Spark-based pipeline was designed to collect, preprocess, and semantically enrich textual data from electronic engineering courses using advanced natural language processing (NLP) techniques. Fine-tuned AI models were integrated into avatars that simulate realistic teaching experiences, aiming to provide scalable and personalized learning opportunities. Evaluation with undergraduate students demonstrated improved accessibility to course content, a measurable reduction in manual preparation time for educators, and increased student engagement compared to static learning materials. While the system strengthens instructional delivery and personalization, it does not replace the essential human roles of mentoring, critical thinking, and creativity. We therefore propose a hybrid educational framework in which AI-driven avatars support repetitive instructional tasks, enabling professors to focus on adaptive guidance and higher-order learning. This work contributes to the growing field of AI in education by demonstrating how avatar-based learning environments can foster inclusivity, interactivity, and efficiency in digital education.