Document type

Extracting data from 401(k) statements

A 401(k) plan reports to every participant each quarter, through whichever recordkeeper the employer hired, and each one brands its statement its own way. A mortgage underwriter reads one to count reserves. Large lenders see thousands a quarter, arriving as portal downloads and phone photographs.

An underwriter needs the account holder, the statement date, the account balance and the contributions behind it. A loan origination system counts the balance toward reserves and reads the contribution rate as evidence of steady income.

Start for Free Talk to Us

Page 1 of a sample 401(k) Statement, with extracted values outlined

27,002Markdown characters

358parse blocks

47fields extracted

The problem

What makes 401(k) statements hard

Scans inside a digital PDF

A plan sponsor's inserts arrive as page images bound into an otherwise digital statement, so some pages carry no text layer.

Employer stock beside cash

Half an account can sit in the employer's own shares, with transfers moving money between that holding and the cash one.

Every recordkeeper brands it

Coloured panels, icon rows and a balance history chart carry the figures that a plainer statement prints in a table.

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 401(k) statement, end to end

About this document: A Voya quarterly statement for a Publix Super Markets 401(k), seven portrait pages. A branded summary and balance history, then holdings split between employer stock and cash, activity, contribution rates and investment elections, and two pages that are scans of the plan sponsor's own inserts. Page 1 is featured. Personal details were replaced before publication.

Processing objective: The objective is to read what a lender checks: whose account it is, when the statement was cut, what it was worth and what was paid in that quarter. The statement date is returned as an ISO date, taken from the period the statement covers.

Demonstration fields: Account holder, Statement date, Account balance, Plan provider, Employee contributions

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 1shown below
Page 3
Page 4
Page 6

Where each value came from

Page 1 of the sample 401(k) Statement, with every featured field outlined

Account holderparticipant_name

JANE Q SAMPLE

The region of the sample 401(k) Statement containing Account holder, with the extracted value outlined

extraction_metadata["participant_name"]

{
  "value": "JANE Q SAMPLE",
  "page": 1,
  "range": {
    "end": 302,
    "start": 289
  },
  "box": {
    "xmin": 0.11412,
    "ymin": 0.14004,
    "xmax": 0.21319,
    "ymax": 0.14998
  }
}

Statement dateperiod_end_date

2023-03-31

The region of the sample 401(k) Statement containing Statement date, with the extracted value outlined

Printed on the page as January 1, 2023 - March 31, 2023, normalized on extraction.

extraction_metadata["period_end_date"]

{
  "value": "2023-03-31",
  "page": 1,
  "range": {
    "end": 685,
    "start": 653
  },
  "box": {
    "xmin": 0.08701,
    "ymin": 0.33625,
    "xmax": 0.31116,
    "ymax": 0.3502
  }
}

Account balanceactivity_summary.closing_value

13,650.02

The region of the sample 401(k) Statement containing Account balance, with the extracted value outlined

extraction_metadata["activity_summary.closing_value"]

{
  "value": 13650.02,
  "page": 1,
  "range": {
    "end": 1190,
    "start": 1180
  },
  "box": {
    "xmin": 0.32013,
    "ymin": 0.56784,
    "xmax": 0.43516,
    "ymax": 0.58726
  }
}

Plan providerrecordkeeper

Voya

The region of the sample 401(k) Statement containing Plan provider, with the extracted value outlined

extraction_metadata["recordkeeper"]

{
  "value": "Voya",
  "page": 1,
  "range": {
    "end": 2669,
    "start": 2664
  },
  "box": {
    "xmin": 0.78494,
    "ymin": 0.94409,
    "xmax": 0.91294,
    "ymax": 0.97429
  }
}

Employee contributionsactivity_summary.employee_contributions

542.08

The region of the sample 401(k) Statement containing Employee contributions, with the extracted value outlined

extraction_metadata["activity_summary.employee_contributions"]

{
  "value": 542.08,
  "page": 1,
  "range": {
    "end": 891,
    "start": 884
  },
  "box": {
    "xmin": 0.32037,
    "ymin": 0.49116,
    "xmax": 0.43522,
    "ymax": 0.5062
  }
}

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.

Compare plans

Capabilities

401(k) Statement processing highlights

Every recordkeeper's layout

Employer stock holdings

Contribution rates by category

Investment elections

Scanned inserts inside PDFs

Fund performance tables

Balance history bar charts

Activity and internal transfers

Vesting and loan balances

Dates normalized to ISO

Bounding box for every value

Zero data retention available

FAQs

Questions, answered

Yes. Parsing works from the rendered page, so a page bound in as an image is read the same way as one with a text layer, and a table on a scanned insert comes back as a table.

Yes. Model the holdings as an array and each comes back with its own units and value, so an employer stock position stays distinct from the cash beside it.

No. A schema names the fields you want. Where they sit on the statement is not part of it, so one schema reads a recordkeeper it has never seen.

Every value comes back with its page, a bounding box and the character range in the parsed Markdown. A date returned in ISO form still points at the period the statement printed, so a reviewer sees both.

A 401(k) statement carries an account holder, a plan number, a balance and a contribution history. A schema can ask for the balances alone. Accounts are available with zero data retention, alongside SOC 2 Type II, GDPR and HIPAA compliance.

Start extracting 401(k) statement data

Reliable, structured outputs with full traceability in minutes