Making bank data accessible
Ledger Tome was built to solve a simple problem: getting transaction data out of PDF bank and credit card statements shouldn't require hours of manual work.
Our Mission
To eliminate manual data entry from financial workflows. We believe accountants, bookkeepers, and finance professionals should spend their time on analysis and advisory, not copying numbers from PDFs into spreadsheets.
Why we built Ledger Tome
From frustration to solution: the journey to automated bank and credit card statement conversion.
The Problem
Accountants and bookkeepers spend countless hours manually entering bank transactions from PDF statements into accounting software. It's tedious, error-prone, and a waste of professional expertise.
The Solution
Extraction reads the document rather than matching it against a stored per-bank layout, so an unfamiliar bank needs no setup. Then the arithmetic on your own statement is used to check the result before you ever see a file.
The Result
Reviewed, reconciled rows exported as CSV, Excel, QBO, QFX, OFX, QIF, MT940, CAMT.053 or JSON, matched to what the destination reads. The retyping goes away; the judgement stays with you.
What we stand for
The principles that guide how we build and operate Ledger Tome.
Show the work, not a score
We publish no accuracy percentage, because we have no benchmark to put behind one. Instead every extracted row is shown against the statement and the running balance is recomputed, so a figure that does not fit its neighbours is flagged before you export.
Say what actually happens to the files
Files are encrypted in transit and at rest, stored in the EU, and deleted after seven days. Converting a document does mean sending it to an AI provider, which is contracted not to train on it; the Privacy Policy names every provider involved.
Reading, not deciding
Retyping a statement is work nobody chose and nobody checks twice. Extraction does the reading; you keep the deciding.
Three steps, no setup
Upload, review, export. No per-bank templates to pick, no mapping to configure before you see anything, and no account required to read what a converter page says it will produce.
How Ledger Tome is run
It is a web application, not an agency and not a service with people in the middle. You upload a statement, a model reads it, you check the rows against the statement, and you export a file. Nobody at Ledger Tome touches your conversion on the way through.
The infrastructure runs on AWS in the EU (Ireland). Uploads and generated files are encrypted at rest and deleted seven days after they are created, or sooner if you delete the conversion. Converting a PDF means sending it to an AI provider for extraction, and those providers are contracted so your documents are not used to train their models. The Security page names them and sets out what is and is not claimed, including the certifications we do not hold.
We do not connect to bank accounts. There is nowhere in the product to enter online banking credentials and no integration that would use them, which is a deliberate limit rather than a feature we have not got to yet.
Support is one address rather than a ticket maze: info@ledgertome.com . If a statement converts badly, sending it to us is the fastest way to get it fixed, and the most useful thing we receive.