Extracting data from title commitments
A title agent issues a commitment ahead of every real estate closing, on its underwriter's version of the ALTA form. The lender's title review team and the escrow officer each work from one per transaction. It arrives as a PDF from the title agent, by email or through a closing portal.
Title review needs the proposed insured, the policy amount, the vested owner, and every Schedule B requirement and exception. Those feed the lender's closing checklist, where requirements are cleared and unacceptable exceptions go back to the agent for removal.
What makes title commitments hard
Schedule B runs for pages
Part II lists each recorded deed of trust, assignment, notice and judgment against the property, each with its own recording number and date.
Requirements and exceptions read alike
Schedule B-I lists what must happen before closing and B-II what the policy will exclude, in numbered lists with the same legal phrasing.
Valid only when complete
Each page of the ALTA form states the commitment is valid only with its schedules and the signature of the company or its issuing agent.
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 commitment for title insurance, end to end
About this document: An ALTA commitment issued by First Integrity Title Company as agent for Westcor Land Title Insurance Company, posted with a property-auction listing. Eleven portrait pages: the signed jacket, the conditions, Schedules A, B-I and B-II, lender notes and privacy notices. Page 1 is featured. Owner names, the property and every recording and order number were replaced with stand-ins, and the wire instructions page was removed.
Processing objective: The objective is to identify the commitment and confirm it was executed: who issued it, on which form, and whether the jacket carries the underwriter's signature and seal. Both attestation checks come back as booleans, each grounded to the mark it was read from.
Demonstration fields: Commitment number, Underwriter, Form edition, Signed, Seal or stamp
The document
4 of 11 pages
Page 1shown below
Page 4
Page 7
Page 8
Where each value came from
commitment_number
{
"value": "FIN-78796415",
"page": 1,
"range": {
"end": 237,
"start": 202
},
"box": {
"xmin": 0.42936,
"ymin": 0.12809,
"xmax": 0.94078,
"ymax": 0.15606
}
}underwriter.name
{
"value": "Westcor Land Title Insurance Company",
"page": 1,
"range": {
"end": 164,
"start": 107
},
"box": {
"xmin": 0.42936,
"ymin": 0.09358,
"xmax": 0.94076,
"ymax": 0.12812
}
}form_edition
{
"value": "ALTA Commitment for Title Insurance (Adapted 08-1-2016)",
"page": 1,
"range": {
"end": 3539,
"start": 3483
},
"box": {
"xmin": 0.58287,
"ymin": 0.93478,
"xmax": 0.93758,
"ymax": 0.94682
}
}attestation.is_signed
{
"value": true,
"page": 1,
"range": {
"end": 2556,
"start": 2548
},
"box": {
"xmin": 0.5825,
"ymin": 0.59061,
"xmax": 0.91864,
"ymax": 0.62833
}
}attestation.has_seal_or_stamp
{
"value": true,
"page": 1,
"range": {
"end": 2473,
"start": 2465
},
"box": {
"xmin": 0.47219,
"ymin": 0.60175,
"xmax": 0.57797,
"ymax": 0.6844
}
}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.
