Invoice Capture – Where to Find Power Automate Flow for Field Validation After AI Model Detection

Power Platform
,
1

Hi everyone,

I’m working with Invoice Capture in Dynamics 365 Finance, and I’m trying to understand the automation flow after the AI Builder model captures fields from a vendor invoice.

Specifically, I would like to know:

  • Is there an existing Power Automate flow that runs after the model detects the invoice data and before it’s inserted into the Captured Invoice or Staging Vendor Invoice Header table?
  • If yes, where can I find this flow (inside a solution or directly in the environment)?
  • Is it possible to add logic to this flow (or create a custom one) to check specific fields after capture—for example, to compare a detected Purchase Order field with a custom field like CustomerPO, and then update or correct it before it moves to the staging tables?

Any guidance or documentation links would be greatly appreciated!

Thanks in advance for your help.

1 Like
2

Hi Ava,
It seems like you have created a document processing model already, You need to have a clarity whether you needs to build a new flow or can use any existing flow template if its already available. Other then that here are answers for your questions :

  1. No, we need to create a custom flow, we have all the AI builder connectors that will help you to read and extract info from the model.
  2. No.
  3. Since the entire flow is custom and completely yours’s, you are free to design it and add logic the way you want.
3

Thank you for your response. Just to clarify we are using the Microsoft Invoice Capture application in Dynamics 365 F&O. The AI model used is a prebuilt one provided by Microsoft, and we have not developed or configured any custom models or flows ourselves. That’s why I’m asking if Invoice Capture includes any built-in Power Automate flows. What I need to know is how exactly does the data from the invoice get populated into the Invoice Capture tables after being captured by the model?

4

Power Automate serves as a bridge between AI models and Dataverse tables, enabling automated data processing and storage. Here’s how the integration works:

  1. Triggering the Flow: You can initiate a flow based on specific events. For instance, when a new email with invoice attachment arrives in Outlook, the flow can be triggered using the “When a new email arrives” connector.
  2. Processing with AI Builder: Once triggered, the flow can utilize AI Builder connectors to process the data. For example, the AI Builder’s document processing model can extract information from email attachments like invoices.
  3. Storing Data in Dataverse: After processing, the extracted data can be stored in Dataverse tables using the “Add a new row” action. This allows for structured storage and easy access to the processed information.

Need any further help you can connect me on LinkedIn - https://www.linkedin.com/in/hardikguptaagt/ or email me at - hardik.gupta@allgrowtech.com