Extracting data from auto declarations pages
Insurers issue a declarations page with every personal and business auto policy and every renewal, each on the carrier's own form. Lenders verify one for every financed vehicle, and underwriters ask for one to prove prior insurance. They arrive by fax and email, or through upload portals.
A lender needs the policy period, each vehicle's VIN, the comprehensive and collision deductibles and the lienholder on the policy. Its insurance tracking system matches those against the loan record, and flags a lapse before the lender places collateral protection insurance.
What makes auto declarations pages hard
Preprinted rows for every coverage
Many carrier forms preprint each coverage on offer, so the purchased lines sit among rows left blank or marked with a bare dollar sign.
Liability limits use two notations
One page prints 100/300/100, three separate caps in thousands of dollars. Another prints a single combined limit for each accident.
Every carrier names the lender differently
One carrier labels the lender a lienholder, another a loss payee or an additional interest, and each prints it in its own place.
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 auto policy declarations page, end to end
About this document: A Business Auto Coverage Declarations page from National Indemnity Company, issued to a limousine and party-bus operator. The Pennsylvania Public Utility Commission posts it as the operator's proof of insurance. It is one portrait page, scanned slightly skewed. Its coverage schedule prices liability and physical damage cover, with a premium against each line.
Processing objective: The objective is to read what the policy buys over its term, and at what price. Liability here is a combined single limit: one figure for bodily injury and property damage in each accident. The estimated total premium sits beneath the charge for endorsements.
Demonstration fields: Policy period from, Policy period to, Liability limit, Liability premium, Estimated total premium
The document
1 page
Page 1shown below
Where each value came from
policy_period.effective_date
{
"value": "10/01/2019",
"page": 1,
"range": {
"end": 545,
"start": 492
},
"box": {
"xmin": 0.07913,
"ymin": 0.26046,
"xmax": 0.49855,
"ymax": 0.27667
}
}policy_period.expiration_date
{
"value": "10/01/2020",
"page": 1,
"range": {
"end": 610,
"start": 597
},
"box": {
"xmin": 0.4842,
"ymin": 0.26218,
"xmax": 0.62784,
"ymax": 0.275
}
}coverages[0].limit
{
"value": "1,000,000 CSL",
"page": 1,
"range": {
"end": 1457,
"start": 1442
},
"box": {
"xmin": 0.45503,
"ymin": 0.40189,
"xmax": 0.83523,
"ymax": 0.41789
}
}coverages[0].premium
{
"value": "5,293",
"page": 1,
"range": {
"end": 1473,
"start": 1466
},
"box": {
"xmin": 0.83521,
"ymin": 0.40187,
"xmax": 0.92014,
"ymax": 0.41774
}
}estimated_total_premium
{
"value": 7173,
"page": 1,
"range": {
"end": 3101,
"start": 3094
},
"box": {
"xmin": 0.83485,
"ymin": 0.63996,
"xmax": 0.92005,
"ymax": 0.6609
}
}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.
