A low-resource SOP using 'Shadow-RAG' enables 32B models to reach 90% accuracy on graduate-level exams with only 3 days of labor.
March 24, 2026
Original Paper
From 50% to Mastery in 3 Days: A Low-Resource SOP for Localizing Graduate-Level AI Tutors via Shadow-RAG
arXiv · 2603.20650
The Takeaway
This provides a highly practical blueprint for creating domain-expert tutors on consumer hardware. It demonstrates that structured reasoning guidance (Shadow-RAG) is more effective than scaling model size for specialized academic tasks.
From the abstract
Deploying high-fidelity AI tutors in schools is often blocked by the Resource Curse -- the need for expensive cloud GPUs and massive data engineering. In this practitioner report, we present a replicable Standard Operating Procedure that breaks this barrier. Using a Vision-Language Model data cleaning strategy and a novel Shadow-RAG architecture, we localized a graduate-level Applied Mathematics tutor using only 3 person-days of non-expert labor and open-weights 32B models deployable on a single