Feature

Your bank statements, classified in one click.

Import a CSV or OFX file from your bank. The reconciliation engine auto-matches every transaction with an issued invoice, a recurring expense, a charges payment or another income line.

Bank import
Auto-reconciliation
CSV · OFX · QIF
3/4 matched
Jun 12
ACME CONSULTING LTD — WIRE
+5,400.00
Matched withInvoice #2026-014 · Acme
Jun 10
CLOUD HOSTING — DD
-89.40
Matched withCloud hosting (recurring)
Jun 08
TAX AUTHORITY — DD
-1,240.00
Matched withQ2 payroll taxes
Jun 04
STRIPE PAYOUT
+320.00
Classify in 1 click

Day-to-day use cases

  • No more typing 80 transactions by hand after every monthly statement.
  • Spot a late client payment or unexpected direct debit instantly.
  • Close monthly bookkeeping in 10 minutes instead of 2 hours.
  • Keep a real-cash trail to calibrate the gap with what your bank sees.

How it works

  1. 1
    You upload your bank's CSV or OFX (most European banks supported).
  2. 2
    The engine parses, detects columns (date, label, amount), and suggests a mapping if needed.
  3. 3
    Each transaction is matched against an issued invoice, a recurring expense, a charges payment, a salary wire or other income.
  4. 4
    You validate transaction by transaction, or in bulk (for example, classify all 4 quarterly URSSAF debits at once).

Details that make the difference

Progressive learning. The engine remembers rules: if you classify "OVH SAS" as "Hosting", next time it's automatic.

Multi-criteria matching. Labels + amounts + dates + tolerances. A €5,400 client payment arriving 47 days after invoice #2026-014 (NET_45 with tolerance) → match.

Special patterns. URSSAF, CIPAV, HMRC, Finanzamt, Stripe are recognized natively.

Bulk mode. If you have 4 identical transactions (same amount, same label), classify all in one click.

Frequently asked questions

Which bank formats are supported?

CSV (exported by most European banks: BNP, HSBC, Barclays, Deutsche Bank, Crédit Agricole, etc.), OFX (industry standard, supported by nearly every bank), QIF (legacy but accepted).

What if the engine doesn't recognize a transaction?

You classify it manually in 2 clicks. The engine remembers and applies the same rule to similar transactions in this import and future ones.

Is my bank data secure?

The file is uploaded over HTTPS, stored encrypted on our NFS PVC, and never shared with any third party. You can delete it at any time.