Administrative overload constitutes an acute structural challenge faced by Indonesian elementary school (SD) teachers implementing the Merdeka Curriculum. This study analyzes the effectiveness of an AI-based system integrating QR Code technology, the Gemini generative AI model, and the Supabase cloud database as a holistic solution, employing a PRISMA-guided Systematic Literature Review (SLR) methodology. Of 312 identified articles, 25 high-quality studies were included following multi-stage screening. Synthesis reveals that: (1) integrated AI systems can reduce assessment administration duration by 60–80%; (2) QR Code technology eliminates student misidentification errors; and (3) generative AI achieves substantial inter-rater agreement (κ ≥ 0.61) for essay response evaluation. These findings underscore the urgency of integrating AI administrative systems specifically designed for Indonesian elementary school contexts to support Merdeka Curriculum implementation.
Copyrights © 2026