Processing invoices containing unstructured data in PDF files and changing them into structured data in the form of CSV files involves two steps. Let's take a look:
- Creating a learning instance: We need to create a learning or training instance by feeding a few sample invoice documents for the AI (IQ Bot) to learn from. Then, we need to review and confirm or correct the mappings so that the AI learns the fields correctly. The IQ Bot is then ready to process invoices that are uploaded to it and convert them into structured CSV files. This can be seen in the following diagram:
- Upload and download from IQ Bot: Once we have the learning instance, we can use a Task Bot to upload the invoices that we'd like to process to the IQ Bot learning instance. The IQ Bot processes the PDF documents and provides structured data in CSV format. We will use another bot to download those CSV files. Here is a sequence diagram showing this process:
Now, let's go...