Extracting data from foreclosure judgments
In judicial foreclosure states, a county court signs a judgment at the end of each case, often on a draft prepared by the lender's law firm. A servicer's default team receives one for every loan foreclosed there, as a scanned PDF from foreclosure counsel or a download from the court docket.
Default servicing needs the case number, the total award with its principal and interest, the fees and the post-judgment rate. The default servicing system records them, sets the credit bid for the sheriff's sale and accrues interest until the sale.
What makes foreclosure judgments hard
The award is a scanned table
The award splits the debt into principal, interest, late charges and fees in a ruled table, often running across a page break.
Each drafter has a format
Florida circuits use a uniform final judgment form, while elsewhere the lender's law firm drafts its own, with its file number in the footer.
Findings written as prose
Interest rates, the deed of trust's recording details and whether a deficiency is awarded sit in numbered paragraphs of findings and orders.
Parse, extract, verify
Parse
Parse turns the document into reading-order Markdown and a hierarchy of blocks (text, tables, figures), and grounds each block to its exact location on the page.
Extract
You define the fields in a JSON schema, and Extract returns their values in a consistent, structured format, each linked back to where it appears in the Markdown.
Verify
Each value links back to where it appears, down to the character range in the Markdown, and the page and bounding box in the source document, so every result can be verified.
One foreclosure judgment, end to end
About this document: A general judgment of foreclosure and sale from the Multnomah County Circuit Court in Oregon, with its certificate of readiness, posted in a sheriff's sale packet. Six scanned portrait pages on line-numbered pleading paper: the caption, numbered findings and orders, the award and fee tables, and the judge's signature. Page 3 is featured. The borrower, defendants, property and file numbers were replaced with stand-ins.
Processing objective: The objective is to lift the money award off a scanned page, one cell at a time. Each amount grounds to its own cell. The prejudgment interest rate is read from inside the row label that states it, beside the per diem that accrues until judgment.
Demonstration fields: Principal, Prejudgment interest, Prejudgment interest rate (%), Late charges, Other costs and fees
The document
4 of 6 pages
Page 1
Page 2
Page 3shown below
Page 4
Where each value came from
amounts_due.principal
{
"value": 352970.28,
"page": 3,
"range": {
"end": 4129,
"start": 4118
},
"box": {
"xmin": 0.53395,
"ymin": 0.58135,
"xmax": 0.71509,
"ymax": 0.59952
}
}amounts_due.prejudgment_interest
{
"value": 59253.99,
"page": 3,
"range": {
"end": 4325,
"start": 4315
},
"box": {
"xmin": 0.71507,
"ymin": 0.59753,
"xmax": 0.85985,
"ymax": 0.66542
}
}amounts_due.prejudgment_interest_rate_percent
{
"value": 3.875,
"page": 3,
"range": {
"end": 4297,
"start": 4169
},
"box": {
"xmin": 0.18015,
"ymin": 0.59952,
"xmax": 0.53407,
"ymax": 0.66852
}
}amounts_due.late_charges
{
"value": 98.6,
"page": 3,
"range": {
"end": 4383,
"start": 4377
},
"box": {
"xmin": 0.53407,
"ymin": 0.66542,
"xmax": 0.71507,
"ymax": 0.68343
}
}amounts_due.other_costs_and_fees
{
"value": 31214.73,
"page": 3,
"range": {
"end": 4475,
"start": 4466
},
"box": {
"xmin": 0.53409,
"ymin": 0.68237,
"xmax": 0.71507,
"ymax": 0.70051
}
}How to try it yourself
Everything on this page is reproducible. The sample, its schema, and the Parse and Extract API outputs are published. You can repeat the same two calls from whichever surface you already build on.
- Claude Skilldescribe the job; the agent writes it
- CLIno code at all
- MCPexpose ADE docs as an agent tool
- REST APIany language
- Pythonlandingai-ade
- TypeScriptlandingai-ade
Documents into data
Accurate
The best accuracy on the difficult documents other tools get wrong, such as messy scans, complicated tables, checkboxes and handwriting.
Auditable
Every answer cites its sources. Values come back with an audit trail (page, bounding box, character range) so a human can review them.
Agentic
Applies multi-step reasoning, and plans before it acts. Handles documents it has never seen before and adapts to variations.
API-first
Composable APIs that drop into any tech stack and can be called from anywhere. Easy to build on and embed in your pipelines and products.
Affordable
Cost-competitive rates with complete price transparency. Spend controls before you run, and every job reports its exact cost.
Enterprise-grade
Built for production and scale from day one. Fully observable and monitored, running securely in our cloud or yours.
Built for high-volume, high-throughput production
- Zero data retention
- Priority rate limits
- Enhanced support
- VPC and on-prem deployment
Available on Team and Enterprise plans.
