Author and systems thinker. I build structurally annotated Arabic research corpora with recorded evidence grades, alternative readings, and negative controls — resources for evaluating language models on classical and high-register Arabic. Qur'anic Structural Research Corpus: 10 datasets, 113 files, 35,932 records, CC BY 4.0. ORCID 0009-0001-7323-1577