Setting Up Power Automate with CoPilot

We began by opening Power Automate and enlisting CoPilot’s help. CoPilot uses prompts to generate automated workflows, but like any partnership, communication is critical. Your “love language” with CoPilot—how you phrase prompts—makes all the difference. Even small tweaks can lead to very different results.

For our workflow, we tested three variations of this prompt:

• Prompt Example 1: “When a new employee is added to a Dataverse table, send an email to that new employee and another to the IT department. Use a lookup with the employee’s name in another Dataverse table to retrieve the required software.”
• Prompt Example 2: “Create a Power Automate flow that triggers when a new employee is added to a Dataverse table. The flow should send a welcome email to the employee and notify IT about the required software.”

Each variation produced slightly different results. After some testing, we selected the workflow that best met our needs.

Refining the Workflow

Once CoPilot generated the flow basics, it was time to dive deeper and fix errors. Initial setup often requires adjustments, so don’t hesitate to fine-tune. In this case, we encountered inaccuracies like invalid table names and parameters. These issues were resolved by manually correcting details like:

• Updating the table name to “New Employees.”
• Adjusting the scope to “Organization.”

Adding Software Lookup and Dynamic Data

The IT notification email needed to include the software required for each employee. We set this up by modifying CoPilot’s suggestions:

1. Changed “Get a Row by ID” to “List Rows” since an employee might need multiple software entries.
2. Inserted a filtering query so only relevant software is included.

To manage the data, we initialized a variable called VAR Software. This variable stores all software entries for the employee, formatted neatly with HTML to enhance readability. CoPilot helped create an Append to String Variable action, which loops through the software entries and adds them to the email dynamically.

Personalizing the Emails

An employee’s first day is exciting, and a warm, personalized welcome email matters. With CoPilot’s help, we added the employee’s first name and a friendly greeting. We also directed them to HR for any questions. On the IT side, the email included a detailed list of software, with each entry showing the software name and required date in the MM/DD/YYYY format.

These details, though small, make automation more human-centered and effective.

Testing and Final Touches

Testing is essential. We created a new employee entry in the Dataverse table to see how the workflow performed. Both emails were sent successfully:

• The employee received a warm welcome.
• The IT department got a clear list of software needs.

For the test phase, we sent both emails to ourselves before finalizing the flow. This ensured everything worked as expected.

Lessons Learned

Working with CoPilot involves trial and error. Sometimes, prompts yield unexpected results, and workflows need manual adjustments. The good news is that every misstep adds to your skills and understanding. If something doesn’t work, ask CoPilot for alternatives, try new phrasing, or manually adjust the flow.

CoPilot isn’t perfect, but it’s a helpful guide that speeds up the learning curve.

Conclusion

Setting up a Power Automate workflow with CoPilot is all about collaboration—between you, the platform, and AI. While CoPilot streamlines much of the process, your role in refining and testing ensures a polished final product.

By automating tasks like sending emails for employee onboarding, you free up time for more strategic work. The workflow we built today is just one example of how you can use AI tools to minimize repetitive tasks and make work more efficient.

Ready to automate your own workflows? Start experimenting with CoPilot and see how it can transform your processes.