Abstract

IDEs (Integrated Development Environments) have become efficient systems which support team work and debugging in coding. However, present IDEs do not help in understanding the core concepts over time and majorly focus on particular tasks. As a result of this, programmers, specifically the ones who are at a beginner or an intermediate level tend to have a messy learning experience, limited grasp over core concepts of programming and get distracted very often. This paper introduces EZ Coder, a hybrid AI-powered mentorship framework designed as a VS Code extension. EZ Coder changes the IDE into an interactive learning environment. Unlike the typical AI assistants that help as passive tools, EZ Coder works as an active mentor by adding three major features into the IDE directly. These features include a personalised roadmap generator which works as an adaptive learning engine, an in-editor code visualizer to increase the understanding of the programs and an AI chatbot which gives context-aware feedback based on Abstract Syntax Tree (AST) analysis of code structure. An AI inference system is being used by the framework. Cloud-based large language models are combined with fast local models for real-time feedback generation and teaching insights. An ongoing, evidencebased learning cycle tracks the programmer’s behavior and updates skill levels using Bayesian reasoning. This cycle prioritizes relevant learning actions without disturbing the workflow. Evaluation of a prototype with programming tasks show that EZ Coder provides extremely accurate and relevant feedback, reduces task completion time and offers much more meaningful guidance than general AI assistants and standard linters.

References

• S. Tipirneni, M. Zhu, and C. K. Reddy, “StructCoder: Structure-aware transformer for code generation,” ACM Transactions on Knowledge Discovery from Data, vol. 37, no. 4, Jan. 2024.
• A. Frommgen et al., “Resolving code review comments with machine learning,” in Proc. IEEE/ACM 46th Int.Conf. Software Engineering: Software Engineering in Practice (ICSE-SEIP), 2024, pp. 52–63.
• A. Mathieu, “Development of animated visualizations of code execution using abstract syntax tree transformations and web technologies,” Ph.D. dissertation, Hochschule fu¨r Angewandte Wissenschaften Hamburg, Hamburg, Germany, 2023.
• Y. Wang, W. Wang, S. Joty, and S. C. H. Hoi, “CodeT5: Identifier-aware unified pre-trained encoder– decoder models for code understanding and generation,” in Proc. Conf. Empirical Methods in Natural Language Processing (EMNLP), 2021, pp. 8696–8708.
• Z. Guan et al., “ContextModule: Improving code completion via repository-level contextual information,” arXiv preprint arXiv:2412.08063, 2024.
• F. Gloeckle et al., “Better and faster large language models via multi-token prediction,” arXiv preprint arXiv:2404.19737, 2024.
• A. Vaswani et al., “Attention is all you need,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 30, 2017, pp. 6000–6010.
• Z. Fan, X. Gao, M. Mirchev, A. Roychoudhury, and S. H. Tan, “Automated repair of programs from large language models,” in Proc. IEEE/ACM 45th Int. Conf. Software Engineering (ICSE), Melbourne, Australia, 2023, pp. 1469–1481.
• A. N. Ramesh, J. K. Singh, and M. P. Satheesh, “AIassisted programming education: Opportunities and challenges,” IEEE Transactions on Learning Technologies, vol. 16, no. 3, pp. 412–425, 2023.
• S. Becker, F. Keller, and T. Fritz, “Reducing cognitive load in software development with context-aware tool support,” ACM Transactions on Software Engineering and Methodology, vol. 32, no. 4, 2023.
• J. Liu, Y. Wang, and S. H. Tan, “Understanding the educational impact of large language models in programming,” in Proc. IEEE/ACM Int. Conf. Software Engineering (ICSE), 2024, pp. 987–998.
• K. T. Chen and P. Brusilovsky, “Adaptive learning paths for programming education using probabilistic skill models,” Computers & Education, vol. 195, 2023.
• M. Mozannar, A. Kapoor, and S. Sontag, “Teaching with AI: Pedagogical implications of generative models,” Communications of the ACM, vol. 66, no. 8, pp. 64–73, 2023.
• A. Ziegler, J. C. Gerlach, and T. Ka¨stner, “IDE-based program analysis for learning-oriented developer tooling,” Empirical Software Engineering, vol. 29, 2024.
• Y. Zhang, Q. Li, and D. Lo, “Code representation learning with abstract syntax trees: A survey,” ACM Computing Surveys, vol. 56, no. 1, 2024.
• R. Karsa, L. Williams, and T. Zimmermann, “Balancing productivity and learning in AI-assisted software development,” IEEE Software, vol. 42, no. 1, pp. 28–35, 2025.