Lightyear customers love the fact that we can automate data extraction and GL coding for the vast majority of their bills, as well as auto-reconciling account statements.
Rarely will there be a document that we can't extract data from automatically, but from time to time, you might need to manually enter an invoice instead. Multi-page documents or PDFs without embedded text layers may present additional challenges for data extraction, requiring manual intervention or the use of specialized tools; see the types of documents Lightyear can receive and process.
When to manually enter a document
In order for data extraction to be possible the document must be a system generated PDF, because real PDFs contain metadata which can be extracted. This metadata is what we use to create a map for data extraction. Lightyear employs Optical Character Recognition (OCR) for PDFs that lack embedded text layers. While OCR enables data extraction for processing, it does not create a selectable text layer, meaning text cannot be copied directly from such PDFs.
If the document doesn't contain any metadata which can be extracted, it will be marked with a keyboard icon: which means that the document will need to be manually typed in.
If you send a map request for a document which doesn't contain any mappable metadata, our mapping team will return the document with a note to let you know this will need to be manually entered, so keep a lookout for any notes we have left you, using the notes icon:
How do I manually enter an invoice?
First, click on the invoice in question. You can click on the Manual option within Doc Data:
This will then allow you to enter all relevant document details and Send to Approvals.
How do I manually enter a statement?
Same as above, first click on the statement to highlight it. In Panel 2 click the dropdown menu More. Then select from the menu that appears, Manually File Statement.
You'll then be asked to complete the following mandatory fields:
Supplier
Statement date
Bill date range
Once you have completed this, it will be filed with the rest of your statements in your Statements Tab.
Best Practices for Invoice Data Extraction
Use the Right File Format: If you encounter issues with text extraction, consider converting the invoice to an Excel file. However, note that:
Smart Extract can process Excel files, but mapping is not supported.
Unstructured layouts in Excel may still prevent reliable extraction across pages.
Manual Entry for Complex Layouts: For invoices with unstructured layouts or missing text layers, manually entering details or using summary maps for totals may be the most effective approach.
Verify OCR Results: Always double-check data extracted via OCR to ensure accuracy, especially for critical billing information.