The tool extracts all text items from every page of your PDF statement using pdfjs-dist, which gives each word's exact position on the page. Items are then clustered into rows by their Y-coordinate: words within 4 points of each other share a row. Within each row, the tool identifies date patterns (dd/mm/yyyy, yyyy-mm-dd, etc.) and monetary amounts, then maps them to the correct columns.
Tips for Best Results
Use the original PDF emailed by your bank — not a scan or screenshot
Multi-page statements work fine; the parser reads all pages
After download, sort by Date in Excel to get chronological order
Use Excel's SUM on the Debit and Credit columns to verify totals match your statement
Frequently Asked Questions
Which banks are supported?
The parser uses Y-coordinate clustering and amount-pattern heuristics rather than bank-specific templates, so it works with statements from most banks — including Chase, Bank of America, Wells Fargo, Barclays, HSBC, Lloyds, and hundreds of others. Statements with very unusual layouts may require manual cleanup in Excel after export.
What columns does the CSV contain?
Date, Description, Debit, Credit, and Balance. The tool identifies amounts by looking for number patterns (e.g. 1,234.56) and assigns them to columns based on position and sign. Debit amounts appear in red and credit amounts in green in the preview table.
What if my statement is a scanned image?
The parser only works on text-based PDFs — those generated by banking software and emailed to you. If your statement was physically scanned (a photo of paper), the text is stored as an image and cannot be extracted without OCR. Run the PDF OCR tool first to make it text-selectable, then re-upload here.
Is my bank statement uploaded to a server?
Absolutely not. The entire parsing process runs in your browser using pdfjs-dist. Your bank statement file never leaves your device and is never transmitted anywhere.
Can I open the CSV in Excel or Google Sheets?
Yes. Click 'Download CSV' and open the file in Excel (File > Open), Google Sheets (File > Import), or any spreadsheet app. The CSV is UTF-8 encoded and quoted correctly to handle commas and special characters in transaction descriptions.