Extracting data from land appraisals
Rural appraisers value farmland and ranches, some on the Uniform Agricultural Appraisal Report and others in narrative reports of their own. Agricultural lenders receive one for every farm real estate loan, and estate planners order them too. Each arrives as a PDF from the appraiser, often dozens of pages long.
Credit review needs the opinion of value, the value per acre and each comparable sale's price, acreage and adjustments. The loan origination system sets the loan-to-value ratio from the opinion, and a review appraiser tests the adjustments against local land sales.
What makes land appraisals hard
Comparables span grid pages
Comparable sales run across several grid pages, each narrow column a chain of per-acre prices and the adjustments applied to them.
Several values, one opinion
Sales comparison, income and cost approaches can each produce a value, repeated on summary pages beside the final opinion.
Prices stated per acre
Prices and adjustments are stated per acre and by land class, such as irrigated cropland or rangeland, then multiplied back to a total.
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 land appraisal, end to end
About this document: Five portrait pages from a Uniform Agricultural Appraisal Report by Warren Land & Livestock, a Montana appraisal firm. It values a grazing ranch of 10,777.89 acres for estate planning. The firm publishes it as a redacted work sample. The pages kept are the summary, two pages of sales grid, the income approach and the reconciliation. Page 2 is featured.
Processing objective: The objective is to follow one comparable sale down its column of the grid, from sale date and acreage to price. The net adjustment per acre then gives the adjusted price per acre, which the appraiser weighs toward a concluded rate. Each value grounds to its own cell in the Sale #1 column.
Demonstration fields: Sale #1 date, Sale #1 acres, Sale #1 price, Net adjustment per acre, Adjusted price per acre
The document
4 of 5 pages
Page 1
Page 2shown below
Page 3
Page 4
Where each value came from
comparable_sales[0].sale_date
{
"value": "10/14",
"page": 2,
"range": {
"end": 5279,
"start": 5274
},
"box": {
"xmin": 0.34309,
"ymin": 0.18184,
"xmax": 0.46069,
"ymax": 0.19701
}
}comparable_sales[0].acres
{
"value": 9781,
"page": 2,
"range": {
"end": 5409,
"start": 5404
},
"box": {
"xmin": 0.34308,
"ymin": 0.19698,
"xmax": 0.46068,
"ymax": 0.21214
}
}comparable_sales[0].sale_price
{
"value": 4050000,
"page": 2,
"range": {
"end": 5533,
"start": 5524
},
"box": {
"xmin": 0.34308,
"ymin": 0.2121,
"xmax": 0.46068,
"ymax": 0.22724
}
}comparable_sales[0].net_adjustments
{
"value": 58,
"page": 2,
"range": {
"end": 8675,
"start": 8673
},
"box": {
"xmin": 0.34296,
"ymin": 0.66646,
"xmax": 0.46052,
"ymax": 0.68171
}
}comparable_sales[0].adjusted_price_per_acre
{
"value": 472,
"page": 2,
"range": {
"end": 8773,
"start": 8770
},
"box": {
"xmin": 0.34295,
"ymin": 0.6817,
"xmax": 0.46052,
"ymax": 0.69696
}
}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. VPC and on-prem deployment require an Enterprise plan.
