When your Unibank statement arrives as a scan or a photo of the page rather than a clean export, the numbers on it are just pixels until they are extracted. Scanned and photographed statements are read with optical character recognition. The engine processes the image, identifies the transaction table and extracts every field it finds. Because no OCR engine reads every digit perfectly, the extraction is verified line by line against the statement’s own printed running balance. A misread digit that breaks the balance is caught immediately and the row is flagged for review, so the file never leaves the tool with a silent error.
Whatever the running balance looks like on your Unibank statement, Convert·Into checks it against the statement's own printed closing figure before you export anything. The running balance printed on the statement is the source of truth. Every row is checked against that figure: a starting balance, each transaction amount and the running total must all agree. If a single cell does not add up, that row is highlighted and held for you to verify. The rest of the statement remains untouched. You get a clean, reconciled file without having to eye-scan columns yourself.
Should a description on your Unibank statement break across multiple lines of the printed page, that line break is not the same thing as a new transaction. Some statements print transaction descriptions that are long enough to wrap across two or three lines. The engine treats these multi-line rows as a single transaction, keeping the description together and the amounts in the correct column. There is no splitting, no shifted rows and no text fragment left behind that might break a formula in the export. The resulting table maps cleanly to Excel columns, CSV fields or QBO fields, ready for import or posting.