SortingHat: Redefining Operating Systems Education with a Tailored Digital Teaching Assistant

SortingHat is an AI-powered digital teaching assistant designed to personalize Operating Systems education using retrieval augmented generation, multi-agent reinforcement learning, and 3D digital human interfaces. The system adapts to individual student learning styles, generates customized exercises, and provides automated grading with personalized feedback to address the traditionally high difficulty of OS courses.

SortingHat addresses a genuine pain point in computer science education: Operating Systems courses have consistently high failure and dropout rates due to their abstract complexity and the heterogeneous backgrounds of students. Traditional lecture-based instruction fails to accommodate varying learning paces and practical needs, leaving many students behind. This new system leverages modern AI infrastructure—specifically RAG frameworks for knowledge retrieval and MARL for adaptive behavior—to create personalized learning pathways that scale beyond what individual instructors can provide.

The broader context reflects growing recognition that one-size-fits-all education cannot succeed in technical domains requiring both conceptual understanding and hands-on practice. Universities increasingly turn to AI tutoring systems to supplement faculty resources and improve outcomes. SortingHat's implementation of a 3D digital human interface powered by large language models represents the convergence of conversational AI with educational technology, enabling empathetic interaction at scale.

The market implications are significant for educational technology vendors and institutions seeking differentiated learning experiences. Higher education institutions facing budget constraints see AI tutors as force multipliers for their teaching staff, potentially reducing dropout rates in foundational CS courses. The system's evaluation pipeline—designed for fair, unbiased grading—addresses a growing concern about algorithmic fairness in education.

Looking ahead, the success of SortingHat will likely depend on deployment outcomes at institutional partners and measurable improvements in student retention and competency. The scalability claims warrant validation against real classroom environments, where technical infrastructure and user adoption present practical challenges beyond the research setting.

• →SortingHat uses RAG and multi-agent reinforcement learning to create personalized OS education paths for students with diverse backgrounds.
• →The system generates adaptive exercises, automated grading, and contextualized feedback to address weak areas in each student's learning.
• →3D digital human interfaces powered by LLMs enable empathetic, scalable tutoring beyond traditional classroom instruction.
• →AI-driven educational assistants represent a growing market for institutions facing resource constraints and high failure rates in technical courses.
• →Success depends on validation through real-world deployment metrics including student retention, competency improvement, and satisfaction.