<![CDATA[Educational inequality continues to affect underrepresented communities, particularly in the context of hybrid learning environments that often fail to account for students’ diverse socio-economic, cognitive, and technological backgrounds. Traditional instructional models rarely provide the necessary flexibility or responsiveness to address learning disparities at scale. This study explores the use of predictive analytics as a tool to personalize hybrid learning trajectories for students in underrepresented communities, aiming to enhance engagement, performance, and retention. The research employed a mixed-methods approach, combining machine learning-based predictive models with qualitative interviews and real-time learning analytics. Conducted across four public schools serving marginalized populations, the study analyzed data from over 300 students to identify risk factors and generate personalized intervention strategies. Results showed that predictive models accurately forecasted student disengagement and academic decline with 85% accuracy, allowing educators to implement timely, targeted instructional responses. Teachers reported improved decision-making and reduced dropout intentions among at-risk students. The study concludes that integrating predictive analytics into hybrid instruction offers a scalable pathway to equity-oriented education, enabling data-driven personalization that supports learners historically excluded from mainstream academic success.]]>
