Exercise - Create a flow to route Helpdesk requests to different mailboxes based on language

Completed

In this exercise, we use a flow template that connects to a preconfigured Language Detection model. This preconfigured Language Detection model is available in environments where sample apps and data are enabled.

  1. Sign in to Power Automate using your organizational account.

  2. Type route emails as per the language detected in the search box and select the Search Icon.

  3. Select the Route emails as per the language detected by AI Builder on new email flow template.

    Screenshot of the Select Route emails as per the language detected by AI Builder on new email flow.

  4. At the bottom of the screen, under This flow will connect to: ensure all of your connections are entered correctly and select Continue. If any of the connections are missing or incorrect, you can't continue.

    Screenshot of all connections entered correctly.

  5. In the first 3 steps of this flow, it's checking a mailbox for new emails, then using the AI Convert email to plain text by inputting 'Body' dynamic content action and then using the AI Builder Language Detection to determine language.

    Screenshot of the first three steps of the flow template.

In this example, we're just using our logged-in user account, but in a real-world scenario this would most likely be a shared Helpdesk mailbox.

  1. This template has a couple of languages predefined, EN and FR. You could add more cases for other languages as well by selecting the + icon.

    Screenshot of the template with predefined languages.

  2. In the Send an email to a specific inbox EN step, select the To field and enter an email address. In this example, I use my mailbox. You can use your email address.

    Screenshot of the Send an email to a specific inbox step.

  3. Repeat the previous step but enter a different email address. In this example, I use our Helpdesk - Europe mailbox.

  4. Select Save.

Any new emails are now routed to the correct Helpdesk mailbox based on the language detected.

The flow should now save successfully.