DOCUMENT INTELLIGENCE

Read any document. Fill any template.

Deterministic document parsing and template filling. Hand Docs a contract, invoice, or tax form and it extracts every field, table, and data point. Then feed data into a template and generate completed documents ready for signature.

WHAT DOCS CAN DO

Six capabilities built for production

Universal Parsing

PDF, DOCX, DOC, scanned images. Docs reads them all and returns structured data you can use.

Field Extraction

Names, dates, amounts, addresses. Every field identified, labeled, and ready for your database.

Table Detection

Complex tables with merged cells, headers, and footnotes. Extracted as clean structured data.

Template Filling

Feed data into any template and get a completed document. PDF or DOCX output, pixel-perfect every time.

Batch Processing

Parse hundreds of documents at once. Queue management, progress tracking, and error handling built in.

Confidence Scores

Every extracted field comes with a confidence score. Flag uncertain extractions for human review.

HOW IT WORKS

Upload a document

PDF, DOCX, or DOC. Docs identifies the format automatically.

Extract data

Fields, tables, and metadata. All structured as clean JSON.

Fill a template

Feed extracted data into a template. Get a completed document.

Send for signature

Hand the filled document directly to Sign. One seamless workflow.

Parse a Documentpython

from masonry import document
result = document.parse(
    file="contract.pdf",
    extract_tables=True
)
print(result.text)
print(result.fields)      # 12 fields found
print(result.tables)      # 3 tables

Fill a Templatepython

filled = document.fill(
    template="nda-template.docx",
    data={
        "company": "Acme Inc",
        "date": "2026-04-21",
        "signer": "Jane Doe"
    }
)
print(filled.download_url)

WORKS WITH

PDFDOCXDOCPython SDKREST APISignFormsMCP