Extracting data from purchase contracts
Most home sales are written on a state REALTOR association's purchase contract, filled in and signed by buyer and seller with addenda attached. Escrow officers and mortgage lenders open a file on every one, receiving the executed contract as an e-signed PDF from the agent or transaction coordinator.
Escrow needs the parties, the property and parcel number, the purchase price, the earnest money and the closing date. The title production system opens the order from those, and the lender sets the loan amount from the price.
What makes purchase and sale agreements hard
A different form in every state
Each state REALTOR association publishes its own purchase contract, with its own section numbering, so the price sits on a different line in each.
Typed values on a printed form
Filled-in values sit in numbered blanks threaded through printed form text, and a date is often split across separate month, day and year blanks.
Later documents change the terms
Counteroffers and addenda can change the price or the closing date the contract sets, so the final terms may sit in a later document.
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 purchase and sale agreement, end to end
About this document: A fully executed Arizona Association of REALTORS residential resale purchase contract, filed publicly on SEC EDGAR as a material contract of a real estate offering. An LLC buys a single-family home from another LLC, all cash. Ten portrait pages with e-signatures and initials on each. Page 1 is featured. Both parties are companies.
Processing objective: The objective is to read the deal from page 1: the parcel being sold and how its price is paid. Earnest money plus the cash due at closing makes up the full price, and each value grounds to its numbered line.
Demonstration fields: Full purchase price, Earnest money, Cash due at closing, Assessor's parcel number, Legal description
The document
4 of 10 pages
Page 1shown below
Page 3
Page 8
Page 10
Where each value came from
price_and_payment.full_purchase_price
{
"value": 845000,
"page": 1,
"range": {
"end": 1366,
"start": 1301
},
"box": {
"xmin": 0.03153,
"ymin": 0.41395,
"xmax": 0.56284,
"ymax": 0.43011
}
}price_and_payment.earnest_money
{
"value": 10000,
"page": 1,
"range": {
"end": 1434,
"start": 1367
},
"box": {
"xmin": 0.05786,
"ymin": 0.43134,
"xmax": 0.96408,
"ymax": 0.44833
}
}price_and_payment.cash_at_closing
{
"value": 835000,
"page": 1,
"range": {
"end": 1473,
"start": 1435
},
"box": {
"xmin": 0.0581,
"ymin": 0.44881,
"xmax": 0.96421,
"ymax": 0.46553
}
}premises.assessor_parcel_number
{
"value": "114-18-0490",
"page": 1,
"range": {
"end": 1195,
"start": 1121
},
"box": {
"xmin": 0.03288,
"ymin": 0.32744,
"xmax": 0.96268,
"ymax": 0.34325
}
}premises.legal_description
{
"value": "Rockcliff Lot 37",
"page": 1,
"range": {
"end": 1292,
"start": 1253
},
"box": {
"xmin": 0.06553,
"ymin": 0.36179,
"xmax": 0.96446,
"ymax": 0.37822
}
}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.
