- Everything you need to know about invoice annotation
Ocerra's data extraction engine is built using AI and machine learning technologies. An advanced AI system requires learning mechanism to understand data from new invoice layouts and get smarter every day.
Invoice annotation allows to recognise data correctly next time you receive an invoice from the same supplier.
In Ocerra, we built-in a user-friendly annotation process that allows us to do so with ease.
Invoice annotation example:
By annotating we are training the system to read data correctly and remember it next time. Annotation only happens for data that is present on the invoice.
Annotation vs manual input
Manually entered data cannot be annotated, thus cannot be remembered by the system. If you manually add something to the invoice, e.g. value or extra line, this is a manual input and needs to be repeated next time.
Account setup annotation process
During account setup we will ask to provide us with sample invoices from your key suppliers. We will prepare and ensure the system is reading data from your key suppliers correctly.
While we keep training the system after you start using it "live" for a few months, we will empower you to use annotation process independently if you ever require it with new invoices in the future.
In Ocerra, there are five types of invoice training & annotation. Click on the link below to learn more about each process:
- Supplier matching
- Value-based annotation or "click-click" (see example above)
- Table-based annotation (annotating columns and lines)
- Single-line annotation (clean table)