Extracting data from trust income documents
A Schedule K-1 (Form 1041) reports a beneficiary's share of estate or trust income on an IRS layout. Tax preparers and lenders may request one when reporting or verifying that income, supplied as a PDF from the trustee or as a scan.
Lenders and preparers need the trust's name and EIN, the beneficiary, and each income line passed through. Tax software carries those onto the beneficiary's return, and a lender documents them as trust income that must continue for three more years.
What makes trust income documents hard
Three parts in two columns
The trust and beneficiary run down the left, and the beneficiary's share of each income line runs down the right, on one page.
Code letters beside amounts
Boxes 11 to 14 pair a code letter with an amount, and one box can hold several coded rows, each tied to its own figure.
Mostly blank by design
A typical beneficiary has two or three of the fourteen boxes filled, with the rest left blank on the form.
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 trust income document, end to end
About this document: The first beneficiary's Schedule K-1 (Form 1041) from an IRS e-file test scenario for a trust return, published so tax software vendors can check their output. One portrait page: the trust and beneficiary on the left, the beneficiary's share of each income line on the right. Page 1 is featured. The beneficiary's number was replaced with a reserved test value.
Processing objective: The objective is to identify the trust and read the income it passed to the beneficiary. The trust's name and EIN come from Part I, and the interest and dividends from the boxes of Part III, each grounded to the box that prints it.
Demonstration fields: Trust name (item B), Trust EIN (item A), Interest income (box 1), Ordinary dividends (box 2a)
The document
1 page
Page 1shown below
Where each value came from
estate_or_trust.name
{
"value": "Cyan Magenta Trust",
"page": 1,
"range": {
"end": 583,
"start": 565
},
"box": {
"xmin": 0.05836,
"ymin": 0.24273,
"xmax": 0.51738,
"ymax": 0.31699
}
}estate_or_trust.ein
{
"value": "00-4012343",
"page": 1,
"range": {
"end": 537,
"start": 526
},
"box": {
"xmin": 0.05769,
"ymin": 0.19736,
"xmax": 0.51738,
"ymax": 0.24198
}
}income.interest_income
{
"value": 8500,
"page": 1,
"range": {
"end": 1291,
"start": 1270
},
"box": {
"xmin": 0.55308,
"ymin": 0.09074,
"xmax": 0.72939,
"ymax": 0.12127
}
}income.ordinary_dividends
{
"value": 189300,
"page": 1,
"range": {
"end": 1388,
"start": 1362
},
"box": {
"xmin": 0.55307,
"ymin": 0.1212,
"xmax": 0.72938,
"ymax": 0.15163
}
}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.
