This article proposes a Question Answering system to ease the work of academic department staff as well as quickly assist students when finding information about academic regulations. The system leverages various language models to address challenges in accessibility and clarity of regulations, ultimately enhancing the student experience. We build an English-Vietnamese automatic question answering system by comparing and evaluating XLM-RoBERTa and other models. These models were fine-tuned and trained on datasets specifically curated from academic regulations at a chosen university. Additionally, we explored different optimizations techniques to improve the model’s understanding of specific information and overall system performance.