Upload a bank statement and your books export. ReconcileIQ matches every line — fuzzy descriptions, date drift, many-to-one splits — and surfaces every exception with a confidence score. CSV, Excel, PDF, OFX, QBO. Any bank, any platform.
Most reconciliation tools give up if your bank's CSV doesn't fit a template. ReconcileIQ auto-detects columns, parses 20+ date formats, handles currency symbols, commas and negative-in-parens, and ships its own PDF converter for the banks that still only offer statements as PDF.
Upload .csv, .xls, .xlsx, .ofx or .qbo. Auto-detection of date, amount and description columns; manual remap if you want to override.
Bank-specific PDF parsing for 17+ UK banks plus US and EU formats. No template setup, no manual fields. Drop a PDF in, get a clean CSV out.
OAuth into your accounting platform. Pick the account, pick the date range, pull the books data — no exports, no copy-paste.
The same engine runs both. Switch modes from the upload screen. Bank rec compares your statement to your accounting record; invoice rec compares outstanding invoices to incoming payments — perfect for AR/AP cleanups and onboarding new clients.
The default. Two datasets — your bank statement and your accounting platform export — matched line by line. Surface unmatched items, suggest fixes, push corrections back to your platform.
Take your AR list and your bank deposits, and match them. Used for AR clean-ups, large client onboardings, and any time you have a list of expected payments that need linking to actual money in.
The engine doesn't just look for exact amounts. It runs every transaction through a sequence of progressively smarter strategies — and surfaces a confidence score so you know whether to trust it or flag it.
Same amount, same date, same direction. Confidence 1.0. Most rows are this — they go through silently.
Same amount, ±3 days for posting drift. Configurable tolerance for slow-clearing transactions, ACH delays, or cross-border timing.
Token similarity on descriptions for cases where the amount is suspiciously close but not exact (FX rounding, fees, wire deductions).
One bank line that splits across multiple book entries — common for batched payouts, payroll runs, daily takings.
Reverse case — multiple bank lines summing to a single book entry. Combined matching with date windows.
For unmatched rows, the engine ranks the 5 most likely candidates with a confidence score. One-click accept or filter, search, refine.
"The matching engine is offloaded to a C++ microservice. A 100,000-line statement reconciles in under thirty seconds. Bigger batches just take a bit longer."
Skip the CSV download dance. ReconcileIQ connects via OAuth to six accounting platforms. Pick the account, pick the date range, run the rec. Discrepancies can be pushed back as one-click fixes.
A note on Xero specifically: Xero's API doesn't allow marking transfers or invoice payments as reconciled — that's a Xero limitation, not ours. The rest of standard transactions reconcile cleanly via the IsReconciled flag. We're upfront about it.
Every rec produces three artefacts: a summary report, the matched pairs, and the exceptions. Export any combination as PDF, CSV, or Excel — for clients, for your file, for the audit.
Executive overview: opening balance, total in, total out, matched count, unmatched count, confidence breakdown. Branded.
Full audit trail of every match: bank line, book line, match confidence, strategy used, any tolerance applied.
Unmatched bank items, unmatched book items, suggestions ranked by likelihood. The to-do list for the close.
Every reconciliation saved to your account. Resume later, compare period-to-period, restore prior state.
Apply filters and export the visible subset, or always pull the full dataset. Useful for client-specific exports.
Once the books are reconciled, jump straight into LedgerIQ for ratio analysis, trend reporting and a board-ready PDF.
A single-month rec for a small business is not interesting. The interesting case is the practice migrating 18 months of data from one platform to another, or the multi-entity finance team with a hundred bank accounts. The engine is built for that.
AES-256 in transit and at rest. GDPR-compliant deletion and consent flows. OAuth tokens stored encrypted, revocable any time. Reconciliation data deleted from temporary processing as soon as the export is generated.
In transit (TLS 1.3) and at rest. All temporary processing files encrypted; deletion auto-runs after export.
Data minimisation by default. Account deletion clears all caches, recs, OAuth tokens, contribution history.
We never see your accounting platform password. OAuth 2.0 tokens scoped to read/write only what we need; revocable any time.
Yes. The CSV importer handles any column format — US date formats (MM/DD/YYYY), USD symbols, parentheses for negatives, and the funny formatting Wells Fargo, Chase, and BofA use. The PDF converter has US bank profiles too. Multi-state sales tax periods, 1099 reconciliation, and multi-entity rec are first-class workflows.
Completely. Upload a CSV from your bank and a CSV from your books, run the rec. No OAuth required. Most one-off recs and migrations work this way. The OAuth integrations are a convenience, not a requirement.
It's bank-specific, not generic OCR. We've parsed the actual statement layouts of the 17+ UK banks listed above, plus US formats. That means amounts come out as numbers, dates as dates, and the multi-line description columns get reconstructed correctly — not the column-soup you get from generic PDF-to-Excel tools.
Two options: upload the PDF and the generic parser tries first (it works for most layouts), or export the statement as CSV from your bank's portal — which most banks support. If you want us to add a specific format, drop us a line.
Tested up to 100,000+ transactions per side in a single rec. The C++ matching engine handles it in under 30 seconds. Bigger jobs run a bit longer; the bottleneck is usually the file upload, not the match.
VAT-aware reconciliation is built in for UK clients. The engine respects VAT codes when matching, surfaces VAT-on-sales versus VAT-on-purchases, and the LedgerIQ handoff includes a VAT analysis module. MTD compliance for VAT submission lives in your accounting platform — we keep your books rec-clean so the submission goes out without surprises.
Yes. The Starter plan is free to use with 5,000 credits per month — enough for a small business or to run a few practice recs. No card required to sign up. See pricing.
Once a transaction is reconciled, CodeIQ codes it. AI-driven account suggestion, VAT classification, invoice OCR, merchant CSV parsing — all auto-posted back to your platform.
CodeIQ features →Once the books are clean, LedgerIQ does the analysis. 40+ modules, ratio trends, board-ready PDFs, anomaly detection — works with any accounting platform's GL export.
LedgerIQ features →ReconcileIQ pays for itself the first time you avoid an evening on a stuck rec. Free to start, no card required.