v1.0 ยท OCR-powered extraction

Turn PDF statements into structured data in seconds.

Upload a bank statement and instantly receive clean JSON with the account holder, account number, statement date, opening and ending balances โ€” ready for reconciliation, KYC, lending, or accounting workflows.

๐Ÿ”’ HTTPS ยท Bearer / X-API-Key ๐Ÿ“„ PDF up to 20MB โšก ~2โ€“5s response ๐Ÿ—‘๏ธ No persistent storage

Restricted access

Sign in with the demo password to open the analyzer.

Don't have access? Contact the administrator.

What you get

PDF โ†’ JSON

Submit a multipart upload of any bank-statement PDF and receive a structured JSON object back, ready to drop into your pipeline.

Smart page selection

Processes the first page plus the last two pages โ€” the parts where account, customer, and balance info live โ€” for fast, accurate results.

Consistent schema

Every response uses the same key names. Missing fields come back as empty strings so your downstream code never crashes on a null.

Key-based auth

Send Authorization: Bearer โ€ฆ or X-API-Key. Keys are validated server-side and never logged in client-facing errors.

Fast turnaround

Typical end-to-end latency of 2โ€“5 seconds per statement, including PDF rendering, OCR, and structured extraction.

Privacy by default

Uploads are processed in a private temp directory and removed after the request. Nothing is retained on disk by default.

How it works

1

Upload a PDF

POST multipart/form-data with a single pdf field (โ‰ค 20MB).

2

Render & analyze

The service rasterizes the relevant pages and runs OCR + vision extraction.

3

Get JSON

Receive a fixed-schema JSON payload you can persist, validate, or display.

Extracted fields

{
  "result": {
    "Bank Name": "Example Bank",
    "Account Holder Name": "Jane Doe",
    "Account Holder Address": "123 Example Street, City",
    "Account Number": "1234 5678 9012",
    "Statement Date": "2024-08-31",
    "Opening Balance": "1,234.56",
    "Ending Balance": "2,345.67"
  }
}