I conduct research at the intersection of NLP, ML, and education for low-resource languages, focusing on Uzbek. My work includes grade-level classification, text-complexity profiling, and book selection for upper-grade students via clustering and logistic regression, with attention to interpretability and classroom readiness. **I also study the mathematical foundations of these methods to implement them rigorously and responsibly, and apply them to improve the Uzbek language ecosystem and the quality of instruction in Uzbek schools.** I curate school corpora, evaluate calibration and fairness, and develop neuro-symbolic, multi-modal intelligent tutoring systems; current efforts explore Logic-Weighted Pooling to improve adaptivity and personalization. Broader interests include educational data mining, readability, and Turkic language resources.