In recent years, traditional English learning assessment methods have struggled to provide timely, personalized feedback that supports learner motivation and engagement. This limitation has generated growing interest in adaptive feedback systems that integrate data-driven evaluation with individual learning trajectories. However, the existing literature lacks large-scale empirical studies that validate such adaptive systems in authentic classroom settings, especially in higher education contexts. To address this gap, this study develops and evaluates an Adaptive Student Evaluation and Feedback System (ASEFS) designed to enhance learners’ English proficiency through continuous, automated feedback. Using data from 250 university students divided into control and experimental groups, we conducted correlation, regression, and Wilcoxon signed-rank analyses. The findings reveal that student engagement, learning strategies, and motivation significantly predict proficiency (R2 = 0.905, p < 0.001), with post-intervention gains of over 10 points in the experimental group. The study contributes a novel, empirically validated framework demonstrating how adaptive evaluation can outperform conventional methods by fostering self-regulated learning and sustained performance improvement.