Extract and filter data from any document.

01 / Core capability

Document Extraction

Send any document — PDF, scanned image, or photo — and tell us exactly what fields to pull out. Invoice numbers, patient names, contract dates, bank balances, passport details. The extraction engine reads the document and returns precisely the data you asked for, as structured JSON.

• —Invoices & purchase orders
• —Medical records & lab reports
• —Contracts & legal agreements
• —Bank statements & payslips
• —Passports & identity documents
• —Any custom document type

02 / Core capability

Data Filtering

You don't always need everything in a document. Define exactly which fields matter to your platform — the API filters out the noise and returns only what you specified. No post-processing, no regex, no manual parsing. The result arrives clean, typed, and ready to store or display.

• —Field-level precision — ask for exactly what you need
• —Null for missing fields, never an error
• —Up to 50 custom fields per request
• —Plain-English field descriptions
• —Consistent JSON schema every time
• —Works across all document types

01

Upload

Send any PDF or image to the API. Invoices, contracts, medical records, bank statements — any document, any format, up to 50 MB.

02

Parse

The document is converted into clean structured text. Tables, line items, multi-column layouts — all preserved accurately.

03

Filter

Accurate AI filtering algorithms locate each field you defined and extract only that data — nothing extra, nothing missing.

04

Integrate

Receive clean JSON via polling or webhook. Plug it directly into your ERP, database, or application with no transformation needed.

One request. Filtered JSON.

A single POST with your document and the fields you want. The API extracts and filters the data, returning only what you asked for as clean JSON.

No schema training. No template setup. No post-processing on your end. Works on invoices, medical documents, legal contracts, and any other document type.

Request — extract & filter

curl -X POST "https://exdocapi.cheapehai.shop/api/v1/extract" \
  -H "Authorization: Bearer YOUR_KEY" \
  -F "file=@invoice.pdf" \
  -F 'extract={
    "document_type": "invoice",
    "fields": {
      "invoice_number": "reference ID",
      "total":          "final amount due",
      "vendor_name":    "seller name",
      "due_date":       "payment deadline"
    }
  }'

Response — filtered data only

{
  "invoice_number": "INV-2024-0042",
  "total":          "₹18,450.00",
  "vendor_name":    "Acme Supplies Ltd.",
  "due_date":       "2024-12-01"
}

Fintech

Invoice & financial data

Automate accounts payable by extracting invoice numbers, totals, vendor details, and line items directly into your ERP or accounting system.

Healthtech

Medical document parsing

Pull patient names, diagnoses, prescriptions, and lab results from consultation notes and reports — structured and ready for your health records system.

HR & Payroll

Contracts & payslips

Extract employee names, designations, salary figures, and joining dates from offer letters and payslips. Onboard faster, with less manual entry.

Logistics

Purchase orders & shipping

Parse PO numbers, supplier names, delivery dates, and line items from purchase orders. Feed your supply chain system automatically.

Banking & KYC

Identity & statements

Extract passport details, account numbers, and transaction summaries from identity documents and bank statements for KYC and onboarding flows.

Legal

Contracts & agreements

Pull party names, effective dates, contract values, and key clauses from NDAs, service agreements, and MOUs — without reading every page.

Any document type

Not just invoices. Medical records, contracts, passports, bank statements, payslips — one API handles all of them with the same interface.

You define the schema

No preset templates to configure. Describe each field in plain English. The extraction engine finds it regardless of document layout or language.

Precise data filtering

Get back only the fields you asked for. No bloated responses, no irrelevant data. Every result is a clean, predictable JSON object.

Async & scalable

Jobs run asynchronously with a priority queue. Poll for status or receive a webhook. Process one document or thousands — the API scales with you.

Reusable schemas

Save field sets as named schemas and reuse them across jobs. Build once, apply everywhere — ideal for platforms processing the same document type repeatedly.

Credit-based, no lock-in

1 credit = 1 page. Buy what you need, use it at your own pace. No subscriptions, no monthly minimums, no vendor lock-in.