AI-Driven Tool Supporting Algorithmic Thinking in Programming Education

Understanding program structure, control flow, and problem decomposition is essential in programming education, but many students find algorithmic thinking challenging. Con-ventional educational approaches frequently prioritize syntax at the expense of logical reasoning and provide minimal per-sonalized feedback. This paper outlines the design and initial assessment of FlowBro, an AI-driven educational tool intended to enhance algorithmic thinking in early programming education. The tool integrates a flowchart visualization, sequential code simulation, and an intelligent assistant that offers contextual hints, debugging tips, and educational support while maintaining a focus on minimalism by not disclosing complete solutions. The system enables students to engage with Python code, follow its execution, and obtain tailored assistance aligned with their specific objectives. A qualitative evaluation with educators in the field revealed the tool's potential advantages, especially in large classes and self-directed learning contexts, while also pointing out challenges concerning the reliability of AI feedback and the necessity for more extensive testing. This study presents a streamlined framework for merging AI with visual learning in programming education, while also suggesting pathways for its future enhancement and empirical testing.