Citation-ready extraction: what auditability actually requires in regulated document workflows
Regulated document workflows need extraction systems that prove their work, not just get it right.
Tomasz Wierzbicki
Section
Features
4 stories in Features.
Extracting tables that span pages, merge cells, and hide their totals in footnotes
Most extraction tools fail silently on complex tables until reconciliation reveals the damage.
Margaret Solloway
Chunking is where most RAG pipelines quietly lose, and document structure should drive the splits
Document structure beats arbitrary token limits for better retrieval results.
Adesola Bankole
Why OCR accuracy benchmarks mislead, and what to measure on your own documents instead
Test OCR tools on your actual documents before deployment, not vendor benchmarks.
Tomasz Wierzbicki