RPA Accounts Payable Invoice Processing Workflow

Hello,
I am new to UiPath. I need a workflow that can:

  1. Open an email attachment and IDENTIFY it is an AP Invoice/Credit Note document
  2. Perform data extraction for key fields like Supplier/invoice number/PO number using OCR/Data Capture
  3. Validate Supplier details against Master Data
  4. Validate PO number against Master Data
  5. Perform Duplicate Check
  6. Route Validation exceptions to an exception queue(s)
  7. Produce XML output file with extracted/indexed data
  8. Save a copy of the invoice document image
  9. Send XML file to a SFTP file location

I have viewed the video tutorial for FULL automation, but I cant recreate the workflow steps. My main objective is to receive a PDF document, extract Key Invoice data, Validate it against business rules, Save a copy of the successfully processed document, produce an XML output file and send the xml file to SFTP location.

Many thanks for all your support.

Luciano

1 Like

Hello Luciano, If you need to extract information from different types of invoices including scanned images, I would suggest to check the Rossum invoice data capture technology.
It can extract data from invoices without any template setup and will return the extracted fields plus metadata (as JSON, XML, CSV). The robot can then manage further invoice workflow including PO validations, duplicate checks, approvals, etc.
Developers’ edition is available here: https://rossum.ai/elis. The API documentation is here: Rossum API Reference. It can be used together with UIPath by using HTTPRequest activity. Sending invoices and receiving extracted data by email (managed by robot) is also possible.

I think from a wider point of view, maybe UiPath stock technology could also suffice, see:

It will not work on OCR files or variable templates, though. There, I concur with TobiasR’s suggestion of using Rossum - Data extraction from invoices - Rossum and UiPath | Rossum is a nice example.

can you send the video link bro…same I need

@balkishan

Is this link what you’re referring to?