Escrow Cases
Methods

How Escrow Cases Works

Escrow Cases turns public court records into reviewed record cards. The public site shows structured metadata and source links; the private corpus stays private.

Public card standard

Every public card needs a caption, court, posture, source link, reviewed date, limitation line, and correction path. The card should answer what happened, what result is shown, what the record hinged on, what documents matter, what the file needed to show, and what the record does not prove.

How to read an Escrow Case record

Publication boundary

Public
129 reviewed court cards, source links, posture labels, topics, limitations, reports, and corrections.
Private
Raw PDFs, OCR, parsed JSON, vendor research output, review notes, embeddings, and operator notes.
Excluded
Company pages, person pages, fraud scores, public raw-document search, and regulatory enforcement rows on this domain.

Source labels

Public pages label records by type, such as court opinion, federal docket, complaint, court order, judgment, court filing, or court index. Vendor and research-tool names stay out of case-card labels.

Current extraction limits

Some intelligence fields are fully extracted; others are inferred from reviewed metadata until the private parsing and extraction pipeline is connected. When an amount has not been extracted from reviewed metadata, the site says so instead of guessing.

Corrections

Corrections are reviewed against public source records before publication. Posture updates, broken links, caption errors, and topic corrections can be submitted from any case page.

Reports · About the data · Submit a correction