Embedding‑Free READ Beats Dense Retrieval on Financial Reports by 43 Points
The READ system replaces neural top‑k retrieval with three deterministic operations – lexical search, structural navigation, and bounded span reads – to audit long, table‑heavy documents. On a 780‑page government financial report it answered 58.8% of verified questions versus 15.7% for dense embeddings, exposing a structural flaw in current RAG pipelines.