Investigación

What we measured, and what moved.

Notes from building cited covenant extraction: how models compare on real agreements, what OCR gets wrong on scans, why the parser runs before the model, how to measure a citation, and how much human review the numbers need. The figures that recur: 97.4 percent citation-exact on the first pass, 99.7 after the second read, and every affected test recomputed inside nine seconds of data landing.

Open-weights models against the frontier on covenant extraction

We ran four classes of model over 60 agreements and 1,840 cited fields. The frontier still leads on raw reading, but the gap sits almost entirely in definition chains and citations. With the deterministic pre-pass the production first pass reaches 97.4 percent, and the second read 99.7.

Read the article

OCR for credit agreements: what actually moves accuracy

Scanned execution versions, two-column schedules, footnoted thresholds and stamped pages. Nine practices that moved our first-pass field accuracy on scans from 88 to 96.9 percent, with the second read closing the rest, and the one metric that hides the damage.

Read the article

Deterministic first, generative second: an architecture for cited extraction

A parser builds the section tree, the definitions index and the cross-reference graph before any model sees the document. The model then answers from candidates and must cite a span the validator can check. Why the order matters, the failure it prevents, and how it recomputes every affected test inside nine seconds of data landing.

Read the article

Right number, wrong clause: measuring citation faithfulness

Field accuracy is the metric everyone reports. It misses the quadrant that fails an audit: the value is correct and the clause it points to is not. That quadrant was 4.6 percent of our fields before the validator and is 1.1 today, inside a 97.4 percent first pass, and it is invisible to a spot check.

Read the article

¿Listo para jubilar la hoja de cálculo de covenants?

Solicite una demo y vea cómo se construye el registro a partir de una indenture o un contrato de crédito real.