Owlgorithm: Supporting Self-Regulated Learning in Competitive Programming through LLM-Driven Reflection

TL;DR

Owlgorithm addresses the gap in SRL-based reflection for competitive programming by using an LLM-driven, two-role (Generator and Reviewer) pipeline to generate Bloom-aligned reflective prompts that adapt to submission outcomes. The approach incorporates a structured five-step workflow and dual-LLM prompt chaining to scaffold self-explanation and debugging, targeting both generalization and diagnostic reflection. An exploratory pilot with teaching staff reveals potential benefits for novice learners, particularly in debugging, while highlighting limitations in feedback accuracy, latency, and interactive usability that demand refinement. The work advances CP education by operationalizing SRL theory within an adaptive, scalable AI tutor that can be integrated into existing contest workflows and LMS environments, with implications for broader reflective learning contexts.

Abstract

We present Owlgorithm, an educational platform that supports Self-Regulated Learning (SRL) in competitive programming (CP) through AI-generated reflective questions. Leveraging GPT-4o, Owlgorithm produces context-aware, metacognitive prompts tailored to individual student submissions. Integrated into a second- and third-year CP course, the system-provided reflective prompts adapted to student outcomes: guiding deeper conceptual insight for correct solutions and structured debugging for partial or failed ones. Our exploratory assessment of student ratings and TA feedback revealed both promising benefits and notable limitations. While many found the generated questions useful for reflection and debugging, concerns were raised about feedback accuracy and classroom usability. These results suggest advantages of LLM-supported reflection for novice programmers, though refinements are needed to ensure reliability and pedagogical value for advanced learners. From our experience, several key insights emerged: GenAI can effectively support structured reflection, but careful prompt design, dynamic adaptation, and usability improvements are critical to realizing their potential in education. We offer specific recommendations for educators using similar tools and outline next steps to enhance Owlgorithm's educational impact. The underlying framework may also generalize to other reflective learning contexts.

Figures (1)

• Figure 1: Workflow of the prompts used in both "All" and "Some or None" Cases Passed. Rounded corners show AI generation, circles show iterative AI refinement, rhombuses show formatting text for processing, sharp-cornered boxes show UI display.