This study presents the design and evaluation of an adaptive intelligent e-learning platform tailored to individual student characteristics such as learning styles (visual, auditory, kinesthetic, textual), engagement, and short-term performance. The system integrates modules for automatic assessment, AI-driven content recommendation, user profiling, and a responsive interface. A mathematical compatibility model using weighted matching was developed, supported by Decision Trees and K-Means clustering for content optimization. Experiments with a simulated dataset (30 students, 50 content items) yielded over 82% accuracy in content matching for students with defined profiles. The platform also effectively identified mismatches for low-activity users. The average learning efficiency score was 0.74 using the proposed metric, EiE_iEi. The platform complies with GDPR standards via secure authentication and encryption. Results confirm that personalized adaptive delivery enhances engagement and outcomes. This research offers a technological foundation for developing intelligent, student-centered learning environments in higher education.