Use case · Bank statements

Bank statement to Excel. Every transaction, reconciled.

Cash-flow underwriting, AML, expense automation, and small-business lending all depend on bank statements being parsed accurately, transaction by transaction. Cogneris turns US bank and credit-card statements — PDF or scan — into a reconciled transaction table in under 3 seconds, ready for Excel, CSV, or your ledger as JSON, with category inference and per-line confidence.

Why bank statements are hard

Every US bank formats statements differently. Chase, Bank of America, Wells Fargo, Capital One, Citi, US Bank, and the regional banks each have their own layout, transaction formatting, and balance presentation. Add credit-card statements and the format space doubles. Cash-flow underwriting needs every transaction parsed; OCR-based parsers hit 70-80% transaction accuracy and miss balances on multi-page statements.

How Cogneris does it

Cogneris parses statements semantically: it identifies the institution, finds the statement period, walks every transaction, normalizes amounts and dates, and reconciles ending balance against the transaction sum. Multi-page statements are stitched automatically; credit-card statements are handled with the same schema as bank statements. Transaction categories (deposit, withdrawal, ACH, wire, fee, interest) are inferred from descriptors.

Sample extraction output

doc_typeBank statement
institutionJPMorgan Chase
account_holderJordan T. Hall
account_number****6789
statement_period2026-03-01 to 2026-03-31
opening_balanceUS$ 18,420.55
total_depositsUS$ 12,450.00
total_withdrawalsUS$ 9,234.18
ending_balanceUS$ 21,636.37
transactions47 transactions parsed
balance_check✓ deposits − withdrawals = balance change
avg_daily_balanceUS$ 19,820.40
nsf_count0
confidence0.98 → auto-approved

Bank statement to Excel in four steps

The extraction API returns JSON. Getting from there to a spreadsheet is a short, repeatable step rather than a separate product:

  1. Upload the statement — POST the PDF or scan to /Document/extraction, attach it to a monitored inbox, or push it through your webhook integration. Multi-month bundles are stitched into one document.
  2. Extract and reconcile — Cogneris identifies the bank, finds the statement period, parses every transaction with a per-line confidence score, and checks that opening balance + deposits − withdrawals = ending balance.
  3. Review what is flagged — transactions below your confidence threshold or failing balance-math come back with the source region they were read from. Add a review step and they wait in a review queue; the rest is auto-approved.
  4. Load it into Excel — the transactions array is a flat table: date, description, amount, running balance, category, confidence. Write it to XLSX or CSV with a few lines of code, or push the JSON straight into your ledger.

Export formats

FormatWhat you getBest for
XLSX (Excel)One row per transaction with typed columns, written from the JSON response as in the sample below.Reconciliation, underwriting review, sharing a statement with a controller.
CSVThe same flat rows as plain text, one file per statement or one per account.Google Sheets, accounting-software imports, data warehouses.
JSONThe full response: transactions, balances, the balance check, confidence, and source citations.REST and webhook integrations, cash-flow models, the audit trail.

From statement to spreadsheet in Python

No client library is needed — the call is a multipart POST with a bearer key, and each extracted cell carries its value and confidence:

import os, pandas as pd, requests

with open("statement.pdf", "rb") as f:
    r = requests.post(os.environ["COGNERIS_BASE_URL"] + "/Document/extraction",
                      headers={"Authorization": "Bearer " + os.environ["COGNERIS_API_KEY"]},
                      files={"file": f})
envelope = r.json()
rows = [{col: cell["value"] for col, cell in tx.items()}
        for tx in envelope["data"]["metadata"]["transactions"]]
pd.DataFrame(rows).to_excel("statement.xlsx", index=False)  # or .to_csv("statement.csv")

What you get out of the box

Every US bank, every credit card

Chase, BoA, Wells, Capital One, Citi, US Bank, regionals, plus Visa / Mastercard / Amex statements.

Transaction-level confidence

Each transaction is returned with a confidence score, source page reference, and bounding box.

Multi-page stitching

30-page statements are stitched into one logical document with transaction continuity preserved.

Balance reconciliation

Opening + deposits − withdrawals = ending. Inconsistencies are flagged before they hit your underwriting model.

Integration patterns

For lenders, fintechs, and expense-automation platforms, Cogneris provides REST APIs and signed webhooks suitable for custom integrations with cash-flow models, lender CRMs, and other downstream systems. The bank statement converter API can also produce CSV-style transaction exports.

Bank statement OCR API evaluation

Benchmark bank statement extraction APIs on transaction row recall, balance reconciliation, duplicate transaction handling, statement-period detection, account-holder normalization, source citations, and whether the API can parse scanned PDFs as well as native PDFs. For lending and underwriting, auto-approval rate matters more than raw OCR accuracy.

Compliance & trust

Bank statements contain account numbers and transaction patterns. The Cogneris platform supports encryption at rest with per-tenant keys where configured. Document and audit retention are tenant-configurable and contract-specific. See our security page for the current control boundaries.

Get started

Pay-per-page pricing means you can start an evaluation today without an annual commit. Most teams ship their first bank-statement extraction into production within a week.

Related extractors

Cogneris extracts dozens of structured document types. The closest neighbors to bank-statement extraction:

For broader context, see the IDP buyer's guide, the 2026 State of Document AI report, or estimate ROI at your volume.

Summarize this page with an AI assistant

Open this page in an assistant with the summary prompt already written. We publish the whole site as plain text at llms-full.txt, so an assistant can read what we actually wrote instead of inferring it.

Assistants summarize, and they get things wrong. Check anything that matters against this page, or ask us directly.