To provide instruction that scales to individual educational needs of the students and addresses the limitations of traditional classroom settings, the Centre for Preparatory Studies (CPS) at Sultan Qaboos University has launched a number of innovative projects, including the one aimed at designing an AI-powered tutor for the English for Business 1 course.
The initiative is conducted under the supervision of the Central Curriculum Committee and reflects the CPS’s commitment to leveraging its internal expertise to enhance teaching and learning.
The initial stage of the project aimed to ensure that the proposed AI-powered tutor system aligns with both pedagogical needs and institutional priorities. Consequently, the project began with a systematic examination of the stakeholders’ perspectives, including their practices of AI tools’ use, perceived concerns and expectations, alongside the review of existing solutions in this area. Based on the analysis of the collected data supported by a comprehensive literature review, on the established pedagogical theories in language learning, the project team developed a structured design for the AI tutor. The design incorporates state-of-the-art AI services to support effective and adaptive learning experiences. It also aligns with contemporary Responsible AI principles, guided by the institutional policies and the AI Risk Management Framework developed by the National Institute of Standards and Technology.
The proposed solution is a web-based platform comprising several integrated modules. A chat module enables students to interact with the tutor by asking questions related to subject content and course administration. This interaction is grounded exclusively in approved course materials, supported by appropriate guardrails, and enriched with scaffolding strategies to guide learning. A practice module allows students to request assessments on specific units and language skills, such as reading and writing, with the tutor providing immediate feedback based on established course rubrics. A dashboard module enables students to monitor their progress and engagement over time.
In addition, a voice module allows students to listen to course materials in audio format, supporting diverse learning preferences.
The project team represents an interdisciplinary collaboration within the CPS, bringing together expertise from its English for Humanities Department (Muhammed Hafeez, Dr. Balaji Srinivasan, Dr. Victoria Tuzlukova) and the Department of Mathematics and IT (Dr. Mohammad Hadra), demonstrating the CPS’s ability to integrate diverse academic strengths in addressing emerging educational needs. The project team has developed a prototype to validate the proposed design and system configuration. The next phase involves translating this design into a Minimum Viable Product (MVP), which will be piloted and systematically evaluated with a group of students in the coming semesters, further reinforcing CPS’s commitment to innovation and continuous improvement in education.
According to Dr. Mohammed Hadra, who leads the AI implementation, the system adopts both Retrieval Augmented Generation (RAG) and advanced RAG approaches to ensure that all tutor responses are rigorously grounded in approved course materials. He further emphasised that the advanced prompt engineering techniques, supported by intelligent request handling mechanisms, are employed to precisely shape the tutor’s behaviour, enabling pedagogically sound interactions that actively promote student engagement and deeper learning.