Document type

Extracting data from master services agreements

Vendors and buyers each draft their own master services agreements, with the pricing in order forms and statements of work attached behind them. Procurement, legal operations and diligence teams review hundreds at a time across a vendor base or a target's customer list. They arrive as e-signed PDFs from contract repositories and data rooms.

Review needs the parties, the term and renewals, the fee cap, the payment terms, the assignment clause and each order line. A contract lifecycle management system tracks renewal dates from those, and accounts payable schedules payments against the order form.

Start for Free Talk to Us

Page 6 of a sample Master Services Agreement, with extracted values outlined

21,489Markdown characters

148parse blocks

75fields extracted

The problem

What makes master services agreements hard

Two documents in one file

The buyer's numbered legal clauses sit in front of the vendor's order form, with its own branding and ruled pricing tables.

Pricing split across periods

An order form prices a prorated first period and then each full year as separate tables, and a table can break across a page.

Key terms inside the prose

A renewal count, a fee cap and an audit window each sit inside a numbered clause, often written in words with figures in brackets.

How it works

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.

Worked example

One master services agreement, end to end

About this document: A Williamson County, Texas master services agreement with Power DMS by NEOGOV, published as an attachment to a public Commissioners Court agenda. Seven portrait pages hold the county's master terms with a scanned signature page, followed by the vendor's two-page order form as Exhibit A. Page 6, the first page of the order form, is featured. The county blacked out the account and order numbers before publishing.

Processing objective: The objective is to capture one subscription line from the order form, along with the prorated total it sits inside. The master terms and the order form have different authors. One schema reads legal prose and a ruled pricing table in the same pass, grounding each value where it is printed.

Demonstration fields: Payment term, Subscription, Users, Prorated period end, Prorated total

The complete example is on GitHubThe sample document, the full extraction schema, the complete parse output and every extracted field.

Open the folder

The document

4 of 7 pages

Page 1
Page 2
Page 5
Page 6shown below

Where each value came from

Page 6 of the sample Master Services Agreement, with every featured field outlined

Payment termorder_form.payment_term

Net 60

The region of the sample Master Services Agreement containing Payment term, with the extracted value outlined

extraction_metadata["order_form.payment_term"]

{
  "value": "Net 60",
  "page": 6,
  "range": {
    "end": 14069,
    "start": 14048
  },
  "box": {
    "xmin": 0.05693,
    "ymin": 0.39083,
    "xmax": 0.23061,
    "ymax": 0.40254
  }
}

Subscriptionorder_form.subscription_periods[0].line_items[2].item

PowerPolicy Professional Subscription

extraction_metadata["order_form.subscription_periods[0].line_items[2].item"]

{
  "value": "PowerPolicy Professional Subscription",
  "page": 6,
  "range": {
    "end": 14984,
    "start": 14947
  },
  "box": {
    "xmin": 0.06033,
    "ymin": 0.64095,
    "xmax": 0.41297,
    "ymax": 0.66342
  }
}

Usersorder_form.subscription_periods[0].line_items[2].quantity

extraction_metadata["order_form.subscription_periods[0].line_items[2].quantity"]

{
  "value": 89,
  "page": 6,
  "range": {
    "end": 15050,
    "start": 15048
  },
  "box": {
    "xmin": 0.75725,
    "ymin": 0.64094,
    "xmax": 0.81881,
    "ymax": 0.66343
  }
}

Prorated period endorder_form.subscription_periods[0].line_items[2].end_date

2024-11-30

Printed on the page as 11/30/2024, normalized on extraction.

extraction_metadata["order_form.subscription_periods[0].line_items[2].end_date"]

{
  "value": "2024-11-30",
  "page": 6,
  "range": {
    "end": 15039,
    "start": 15029
  },
  "box": {
    "xmin": 0.65146,
    "ymin": 0.64094,
    "xmax": 0.75725,
    "ymax": 0.66343
  }
}

Prorated totalorder_form.subscription_periods[0].period_total_usd

1,260.69

extraction_metadata["order_form.subscription_periods[0].period_total_usd"]

{
  "value": 1260.69,
  "page": 6,
  "range": {
    "end": 15527,
    "start": 15518
  },
  "box": {
    "xmin": 0.81873,
    "ymin": 0.69347,
    "xmax": 0.94261,
    "ymax": 0.71606
  }
}

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

Run this sample yourself This document on GitHub

Why LandingAI?

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.

Compare plans

Capabilities

Master Services Agreement processing highlights

Parties and effective dates

Initial term and renewals

Not-to-exceed amounts

Payment terms

Order form line items

Subscription periods and quantities

Assignment and audit clauses

Governing law and venue

E-signature blocks

Scanned pages in digital files

Amounts in words and figures

Bounding box for every value

FAQs

Questions, answered

Yes. Extraction reads the whole parsed file. A single schema pulls contract terms from the clauses and pricing from the order form, with each value grounded to its own page.

Yes. Define subscription periods as an array, each with its own line items and total. Every period comes back with its rows, including a table that continues onto the next page.

Dates come back in ISO form, year first. Grounding still points at the date as the vendor printed it, so a reviewer can compare the normalized value with the original string on the page.

Signature blocks parse as attestation blocks, labeled SIGNED, E-SIGNED, STAMPED or SEALED. Printed names and timestamps inside them are transcribed. Side-by-side signature columns stay separate.

Every extracted value carries the character range it was quoted from in the parsed Markdown. The Ground API maps that range to a page and a bounding box, so a reviewer sees each term in the clause or cell that states it.

Start extracting master services agreement data

Reliable, structured outputs with full traceability in minutes