Extracting data from stock purchase agreements
Each law firm drafts its stock purchase agreements with its own section numbering and defined terms. Lenders, insurers and regulators review the executed copy, often a scanned PDF from the closing set or a filing exhibit. A corporate archive holds one for every acquisition closed.
Reviewers extract the seller, the buyer, the target company, the shares sold, the purchase price and the closing terms. A contract management system stores them as the deal record, and the termination fee and outside date set the deadlines tracked after signing.
What makes stock purchase agreements hard
Values defined mid-sentence
Each party and figure is introduced inside a sentence with a defined term, such as (the "Purchase Price"), and later sections use only that term.
Terms spread across articles
Price adjustments, indemnity caps and termination rights sit in separate articles dozens of pages apart, each pointing back to the definitions article.
Agreements arrive inside filings
A purchase agreement often travels as an exhibit to an SEC report or regulatory application, whose cover pages restate the price in different words.
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 stock purchase agreement, end to end
About this document: An executed stock purchase agreement under which a holding company buys every share of the parent of a water and wastewater utility group. It was filed as an exhibit to a joint application with the Kentucky Public Service Commission. The 78-page filing is a raster scan, with the agreement on pages 31 to 75. Page 32, its first page of terms, is featured.
Processing objective: The objective is to capture the whole transaction from the agreement's first page of terms, where the recitals and Section 1.02 set it out. Each value grounds to its printed line on the scan, within the contract itself.
Demonstration fields: Seller, Buyer, Company whose stock is sold, Shares sold, Purchase price
The document
4 of 78 pages
Page 31
Page 32shown below
Page 33
Page 63
Where each value came from
seller.name
{
"value": "Nuon Global Solutions USA B.V.",
"page": 32,
"range": {
"end": 40915,
"start": 40823
},
"box": {
"xmin": 0.1399,
"ymin": 0.16186,
"xmax": 0.86526,
"ymax": 0.18004
}
}buyer.name
{
"value": "Hydro Star, LLC",
"page": 32,
"range": {
"end": 41009,
"start": 40916
},
"box": {
"xmin": 0.14106,
"ymin": 0.17841,
"xmax": 0.86507,
"ymax": 0.19709
}
}company.name
{
"value": "Nuon Global Solutions USA Inc.",
"page": 32,
"range": {
"end": 41240,
"start": 41149
},
"box": {
"xmin": 0.14191,
"ymin": 0.27692,
"xmax": 0.86596,
"ymax": 0.29503
}
}stock.number_of_shares
{
"value": 1000,
"page": 32,
"range": {
"end": 41148,
"start": 41070
},
"box": {
"xmin": 0.1975,
"ymin": 0.25951,
"xmax": 0.86644,
"ymax": 0.2779
}
}purchase_price
{
"value": 192500000,
"page": 32,
"range": {
"end": 42654,
"start": 42560
},
"box": {
"xmin": 0.14649,
"ymin": 0.70523,
"xmax": 0.87003,
"ymax": 0.72522
}
}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.
