When your Bankinter 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. Even a statement that arrives as a scan or a photograph is read cleanly by the engine. OCR extracts every transaction row, and the running‑balance check catches anything the scan may have misread: a 3 that became an 8, a decimal point that disappeared. The extraction works the same way whether the document was exported as a clean digital file or photographed with a phone, because every row is verified against the statement’s own printed closing figure. A row that does not add up is flagged for review, so no error slips into the ledger uncaught.
If a transaction description on your Bankinter statement is long enough to wrap across two or three lines, it still needs to land in a single row once converted. When transaction descriptions wrap across multiple lines, the engine keeps the record whole. It reads the layout pattern rather than assuming one line equals one row, so a long narrative memo that spills into a second line is still assigned to its correct transaction. The extracted table maps each field to its proper column: date, description, amount and the running balance. Nothing is split across rows, and nothing is merged into a neighbouring cell.
If your Bankinter statement runs to several pages, uploading it starts the same layout detection Convert·Into applies to any statement document, page by page. A statement document that runs several pages long is handled as one continuous set of transactions. The engine reads across page breaks without duplicating the header row or dropping the first transaction after the fold. The same balance verification runs from the opening figure on page one through to the closing figure on the final page. Export the complete reconciled table to Excel, CSV or QBO, with all columns mapped and no manual reassembly required.