Extracting data from D&O insurance policies
Insurers issue directors and officers policies on their own forms, with a declarations page followed by general conditions, a coverage part and endorsements. Brokers and risk managers review each policy at renewal, and a broker's management liability book can run to hundreds. Policies arrive as PDFs from carrier portals or by email.
Policy checking needs the limit of liability, the retention, the premium, the policy period and the extended reporting terms. A broker's agency management system compares them with the expiring policy and the binder, and flags any change before the policy goes to the client.
What makes D&O insurance policies hard
Declarations amid long conditions
The limit, retention and premium appear once on the declarations, while terms such as Limit of Liability recur dozens of times in the conditions.
Key terms are stated in prose
Whether the policy is claims-made, and whether defense costs erode the limit, are stated in capitalized warnings and in the conditions.
Endorsements amend the base form
Each policy attaches its own endorsements, which add entities, change definitions or amend exclusions in conditions printed pages earlier.
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 D&O insurance policy, end to end
About this document: An issued Travelers management liability policy with Mutual Fund Liability coverage, filed publicly with the SEC by a registered fund trust. 26 portrait pages: the declarations, terrorism endorsements, general terms and conditions, the coverage part and further endorsements. It covers the trust's directors, trustees and officers. Page 2 is featured. The filing's cover letter and board resolutions were dropped, since they sit outside the policy.
Processing objective: The objective is to read the policy's economics from the declarations. They show how much cover was bought, the retention paid on each claim and the annual premium. The extended reporting terms price 12 more months of claim reporting after the policy ends.
Demonstration fields: Limit of liability, Retention per claim, Policy premium, Extended reporting premium (%), Extended reporting months
The document
4 of 26 pages
Page 1
Page 2shown below
Page 6
Page 17
Where each value came from
coverage_details.limit_of_liability
{
"value": 1000000,
"page": 2,
"range": {
"end": 1692,
"start": 1665
},
"box": {
"xmin": 0.05862,
"ymin": 0.32826,
"xmax": 0.38363,
"ymax": 0.34523
}
}coverage_details.retention
{
"value": 75000,
"page": 2,
"range": {
"end": 2073,
"start": 2045
},
"box": {
"xmin": 0.05898,
"ymin": 0.51184,
"xmax": 0.39204,
"ymax": 0.52809
}
}policy_premium
{
"value": 14366,
"page": 2,
"range": {
"end": 2251,
"start": 2181
},
"box": {
"xmin": 0.05896,
"ymin": 0.60539,
"xmax": 0.74971,
"ymax": 0.62158
}
}extended_reporting_period.additional_premium_percentage
{
"value": 125,
"page": 2,
"range": {
"end": 2349,
"start": 2312
},
"box": {
"xmin": 0.05988,
"ymin": 0.68207,
"xmax": 0.3571,
"ymax": 0.69743
}
}extended_reporting_period.additional_months
{
"value": 12,
"page": 2,
"range": {
"end": 2372,
"start": 2350
},
"box": {
"xmin": 0.06049,
"ymin": 0.69972,
"xmax": 0.3292,
"ymax": 0.71474
}
}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.
