01CORE
Fuzzy matching
Numbers, dates, references and free-text descriptions compared with tolerance you set once and forget.
EST. 2026 · BANK ↔ LEDGER RECONCILIATION
Your bank says one thing. Your books say another. ReconcileBot reads both sides, tests every pair, and points at the money that disagrees.
12,480 ROWS / RUN99.2% MATCH RATE~4 MIN AVERAGESCROLL
STEP 01 — INTAKE
Two files. Nothing in common but the money.
Drag in a bank export and a ledger dump. Different columns, different dates, different spellings of the same vendor. Fine.
STEP 02 — MATCH
Every pair tested, not eyeballed.
Amount, date, reference, description — exact where it must be, fuzzy where it should be. 1,196 candidate pairs, resolved in one pass.
STEP 03 — SURFACE
The 3% that disagrees, ranked by cost.
Everything green settles quietly. What's left is amber and red — the only rows a human should ever look at.
— MATCHED 0— MISMATCH 0— MISSING 0
03 / THE NUMBERS THAT MATTER TO A CONTROLLER
99.2%
MATCH ACCURACY
Across 340 test reconciliations on real-world bank and ledger exports.
95%
LESS MANUAL ENTRY
The rows you actually touch drop from thousands to the few hundred that disagree.
4 MIN
AVERAGE RUN TIME
From drag-and-drop to exportable report, including anomaly scoring.
12,480
ROWS PER RUN
Per file pair, on the free tier, without asking anyone for permission.
04 / FIVE MOVES, START TO EXPORT
Upload files
Drag a CSV out of any accounting system. No template, no cleanup, no column renaming ritual.
01 — 05
01Upload files0:20
02Confirm the keys0:40
03Run the match2:10
04Review the exceptions3:30
05Export the report4:00
05 / YOUR MONTH-END, SHORTER
Close the books
before lunch.
FREE FOR ACCOUNTANTS & SMALL BUSINESSES · NO CARD