Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces EduMate.AI, a generative AI tutoring system designed specifically for beginner C++ learners, addressing the challenges of syntax, logic, and debugging through personalized support. By employing a Model-View-Controller (MVC) architecture and constrained prompt orchestration, EduMate.AI shifts the focus from merely providing answers to enhancing learners' conceptual understanding and reasoning skills. The key result demonstrates that this approach significantly improves the overall learning experience, fostering deeper engagement with programming concepts.
EduMate.AI transforms the role of generative AI from answer delivery to structured tutoring, enhancing programming skills in beginners.
Learning programming language is difficult for many beginners, not only because of syntax, logic deployment, and debugging issues, but also due to the lack of immediate personalized support. General-purpose LLM tools can help but tend to promote answer dependency by generating complete answers without enforcing reasoning. This paper presents EduMate.AI, a logic-based generative AI tutoring system for beginner C++ learners. The system is implemented using a Model-View-Controller (MVC) architecture, integrating a large language model via constrained prompt orchestration. EduMate.AI aims to improve learners’ conceptual understanding, reasoning ability, and learning experience by shifting generative AI from answer-delivery to structured tutoring support.