AI-Driven Personalization in Education: Designing Adaptive Learning and Mentorship Systems

Abstract

Artificial intelligence is rapidly transforming educational paradigms, enabling the design of systems that adapt to individual learner needs, optimize mentorship quality, and enhance academic outcomes at scale. This study investigates the design, implementation, and evaluation of AI-driven personalization frameworks in educational settings, focusing on adaptive learning systems and intelligent mentorship platforms. Using a mixed-methods approach with survey data from 620 learners and 180 educators collected between 2022 and 2025, the study integrates Canonical Correlation Analysis (CCA) to examine the relationship between AI system design parameters and learning outcomes, alongside descriptive and inferential statistics. Results indicate that AI-driven personalization significantly improves engagement, knowledge retention, and learner satisfaction, with mentorship systems showing the highest impact on self-regulated learning. The study identifies critical design variables including feedback latency, content adaptability, and mentor-AI collaboration that predict positive outcomes. Findings suggest that ethically governed, pedagogically informed AI systems hold considerable promise for equitable and scalable educational transformation.