Extracting data from employment verification letters
An employer writes a verification letter whenever someone asks it to confirm that a person works there. The format is entirely the employer's own, drafted by each HR team on company letterhead. Lenders, landlords and visa officers collect them by the thousand, as emailed PDFs and scans of the signed page.
A verifier needs the employer, the employee's start date and title, and the signature that attests to them. A background screening system files the result against the applicant, and an underwriter treats the letter as the employment evidence in the file.
What makes employment verification letters hard
The signature is the payload
A verification letter carries its weight through the signature, which sits on the page as an image among the printed text.
Two dates, different meanings
The letter is dated at the top and the signature carries its own timestamp underneath, and a verifier needs both.
The issuer is the letterhead
Who vouches for the employee is carried by the logo and letterhead block, where a reader recognizes the company by sight.
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 employment verification letter, end to end
About this document: An employment certificate from Deloitte Consulting India. One portrait page on company letterhead, confirming a start date and designation in two justified sentences, with a residential address, an office address and a signature block. This is a real certificate. The employee's name, ID and home address were replaced before publication; the signatory's signature is kept.
Processing objective: The objective is to establish the whole attestation: who issued the letter, on what date, for which role, and whether it carries a signature. The signature field grounds to the signature image itself, so the crop shows the mark the boolean was read from.
Demonstration fields: Letter date, Employer (letterhead), Designation, Signed, Signature date
The document
1 page
Page 1shown below
Where each value came from
letter_date
{
"value": "April 25, 2022",
"page": 1,
"range": {
"end": 234,
"start": 220
},
"box": {
"xmin": 0.08993,
"ymin": 0.22735,
"xmax": 0.19751,
"ymax": 0.24161
}
}employer.name
{
"value": "Deloitte Consulting India Private Limited",
"page": 1,
"range": {
"end": 57,
"start": 12
},
"box": {
"xmin": 0.5888,
"ymin": 0.06824,
"xmax": 0.84435,
"ymax": 0.08026
}
}employee.designation
{
"value": "XIN-DC SENIOR CONSULTANT",
"page": 1,
"range": {
"end": 460,
"start": 363
},
"box": {
"xmin": 0.08843,
"ymin": 0.34496,
"xmax": 0.80922,
"ymax": 0.35956
}
}signatures[0].is_signed
{
"value": true,
"page": 1,
"range": {
"end": 1077,
"start": 1069
},
"box": {
"xmin": 0.08599,
"ymin": 0.68161,
"xmax": 0.30896,
"ymax": 0.75164
}
}signatures[0].signature_date
{
"value": "04.25.2022; 19:22:46 IST",
"page": 1,
"range": {
"end": 1130,
"start": 1100
},
"box": {
"xmin": 0.08685,
"ymin": 0.7531,
"xmax": 0.24983,
"ymax": 0.76366
}
}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.
