Extracting data from rent rolls
Property managers export rent rolls from their management software, each in its own columns and order. Multifamily lenders, buyers and appraisers collect a current one for every property they underwrite, as a PDF from the owner, the broker's offering memorandum or a data room.
Underwriting needs each unit's type, status, square footage and rent, along with occupancy and total rent. The underwriting model builds gross potential rent and vacancy from those, and sizes the loan against the resulting net operating income.
What makes rent rolls hard
Every row looks alike
A rent roll runs one row per unit, and each value has to stay with its unit through dozens or hundreds of near-identical rows.
Two values in one cell
Columns often combine fields, such as bedrooms and bathrooms printed together as 3/2.00, which a schema needs as separate numbers.
Each software exports its own
Property management systems each export different columns in a different order, from a bare unit list to full lease and deposit detail.
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 rent roll, end to end
About this document: The rent roll page from a broker's offering memorandum for a 25-unit apartment property in Washington State, published on the broker's site. One landscape page: a header with the property, a table of unit rows and a totals row. Page 1 is featured. The page lists units and rents only.
Processing objective: The objective is to read the totals row as one summary of the property: how many units, how full, how much space and how much rent. Each total grounds to its own cell, and the unit rows above it add up to the rent and area printed there.
Demonstration fields: Total units, Occupancy, Total square feet, Total monthly rent
The document
1 page
Page 1shown below
Where each value came from
totals.total_units
{
"value": 25,
"page": 1,
"range": {
"end": 2690,
"start": 2682
},
"box": {
"xmin": 0.16161,
"ymin": 0.83565,
"xmax": 0.27493,
"ymax": 0.86916
}
}totals.occupancy_percent
{
"value": 100,
"page": 1,
"range": {
"end": 2716,
"start": 2708
},
"box": {
"xmin": 0.38194,
"ymin": 0.83557,
"xmax": 0.50196,
"ymax": 0.86916
}
}totals.total_square_feet
{
"value": 23760,
"page": 1,
"range": {
"end": 2731,
"start": 2725
},
"box": {
"xmin": 0.50195,
"ymin": 0.83553,
"xmax": 0.60397,
"ymax": 0.86915
}
}totals.total_monthly_rent
{
"value": 22280,
"page": 1,
"range": {
"end": 2747,
"start": 2740
},
"box": {
"xmin": 0.60395,
"ymin": 0.8355,
"xmax": 0.69761,
"ymax": 0.86915
}
}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.
