Every cloud reconciliation tool holds your bank data on its servers. On-device matching keeps it on your computer. Here is how that works and how to check it.
Cloud reconciliation and AP platforms store invoice, vendor and bank data in the vendor's cloud. On-device matching runs in the browser on your own computer: files are read locally, matched locally and never sent. A page security rule (connect-src 'none') blocks every outside request, which you can verify by turning off Wi-Fi after the page loads.
| Cloud reconciliation | On-device matching | |
|---|---|---|
| Where the data goes | Uploaded to the vendor's servers | Stays on your computer |
| Connects to your bank | Often, by feed | No; you load statement files |
| Set-up | Account, connections, implementation | Open the page, load two files |
| Pricing | Subscription per user, client or transaction | Once-off licence, no subscription (AppIQ AI Bank Match Pro, US$1,997) |
| Works offline | No | Yes, after the page has loaded |
| Suits | High volumes, many entities, integrated workflows | Files from systems that do not talk to each other; data that may not leave the company; bookkeepers with many clients |
Bank statements and ledgers are among the most sensitive files a business has. Handing them to a vendor means trusting the vendor's security, its staff and its continuity. That is not an argument against the cloud for every job; it is a reason to decide deliberately which data leaves the building.
AppIQ AI Bank Match Pro is a web page. When it loads, the browser downloads the page, the matching code and a small AI model from appiqfinance.com. From that point the page needs nothing else: you choose the bank statement file and the ledger export on your computer, the page reads them in memory, runs rule-based and AI matching, and shows the results. Nothing is sent. The reconciliation report is built in your browser and saved to your computer.
The guarantee is technical, not a promise. The page is served with a security rule, connect-src 'none', which instructs the browser to block every outside request from that page. You can confirm it yourself: open the page, turn your Wi-Fi off, load your files and match. If it works with the network off, the data stayed put.
Bank statement exports as CSV, Excel, OFX/QFX, CAMT.053 and MT940, and ledger or cashbook exports as CSV or Excel. See the product page for the current list and limits.
If your ledger already holds both sides and offers a bank feed, use it; TechRepublic's comparison of bank reconciliation software covers the built-in tools. If you match tens of thousands of lines a month across entities with approval workflows, a platform earns its price. On-device matching is for the gap between those: files, privacy, speed and no subscription.
UK organisations handling personal data in bank lines should record where processing happens for their GDPR records; on-device processing simplifies that answer. US and Australian teams face the same question under their own privacy and client-confidentiality rules; check your obligations.
AI software that matches your bank statement to your ledger on your own computer and builds the reconciliation statement. $1,997 once-off; see it on the built-in example first.
The Bank Reconciliation Template reconciles up to five bank accounts across eight transaction categories, with an allocation check that confirms every transaction is coded before the reconciliation counts as clean. Excel, no macros, one-time price.
Yes. AppIQ AI Bank Match Pro runs in the browser on your own computer; files are read and matched locally and the page is blocked from sending data anywhere. Turn off Wi-Fi after loading to verify.
No. You export a statement file from your bank and a ledger file from your accounting system and load both; nothing connects to a bank account.
Accuracy depends on the matching rules and model, not on where it runs. AppIQ publishes its test method and results on the product page; run your own files through the free tier and compare.
Last updated 06 November 2026. General guidance for finance teams; check it against your own policies, systems and local rules.