We present the Term Paper Recommendation System (TPRS), a hybrid AI-powered platform designed to scaffold the development of academic term papers in distance education. Tailored for students in a Bachelor of Arts program in Culture and Social Sciences, TPRS dynamically integrates large language models (LLMs), expert- and knowledge-based recommendation engines, and sentiment-driven routing to provide personalized formative feedback. A multi-shot prompting technique simulates high-fidelity tutoring interactions, trained on real supervision logs. Our system prioritizes transparency, student autonomy, and pedagogical alignment by combining structured validation, explainable recommendations, and relevance-based literature suggestions. A pilot deployment involving 18 students showed statistically significant improvements in submission quality (Hedge’s g = 0.44, p < .05) and positive user reception across accuracy, usability, and trust dimensions, evaluated using the CRS-Que framework. This work contributes a modular, pedagogically informed approach to AI-supported academic writing, offering a promising direction for scalable, inquiry-driven support systems in higher education.