Exercise 1, Task 4: Use the Copilot Analyst agent to analyze customer support data

Lamna Healthcare Company has been assisting Trey Research, a medical research organization running multi‑site clinical trials. Over the past 90 days, Lamna’s Support team recorded a surge of issues reported by Trey Research—ranging from appointment sync delays and Electronic Health Record (EHR) synchronization update lags to intermittent sync failures and slow module loads in Scheduling and EHR‑Sync components. Lamna’s Leadership team needs clear, data‑driven insight into what’s happening so they can prioritize fixes, coordinate with Engineering and Product, and prevent disruption to trial operations.

As Lamna’s Customer Service Manager, you plan to use Copilot’s Analyst agent to analyze Trey Research’s dataset, identify patterns in Trey Research’s 90‑day case history, and generate the most meaningful visuals.

Perform the following steps to complete this task:

  1. Select the following link to download the TreyResearch_Support_Cases.xlsx file and then store it in your OneDrive account so that Copilot can access it from your tenant.

  2. Open the TreyResearch_Support_Cases.xlsx spreadsheet and review the results. You might have to expand each column width to see the full extent of the answers. Close the spreadsheet once you’re done.

  3. In your Microsoft Edge browser, go to the Microsoft 365 home page and select the Analyst agent in the navigation pane.

  4. In the Analyst agent, attach the TreyResearch_Support_Cases.xlsx spreadsheet. Then ask the agent to analyze the attached dataset and produce a concise briefing that includes: top issue categories, severity mix, modules most impacted, escalation rate, and resolution-time distribution.

  5. Review the results. Everything looks good, so ask the agent to test correlations between Severity and escalation rate, and between Module and average resolution hours. Generate the three visuals that would communicate these findings most clearly.

  6. Review the visuals. The agent might provide a suggested prompt asking if you want it to combine the three visuals into one dashboard-style summary image for easy presentation. Select and submit this prompt. If the agent doesn’t display this prompt in your lab, then manually enter this request.

  7. You really like this one-page summary. If the agent provides a suggested prompt to add annotations and key insights directly on this dashboard, then select and submit this prompt. If the agent doesn’t display this prompt in your lab, then manually enter this request.

  8. Review the updated summary that includes the annotations and key insights. If the agent provides a suggested prompt to create a presentation-ready version with a summary text panel alongside the charts, then select and submit this prompt. If the agent doesn’t display this prompt in your lab, then manually enter this request.

  9. After the agent generates this presentation-ready version, it might ask if you want it to export this image as a PowerPoint slide (PPTX) for direct use in presentations. In our testing, it asked if we wanted it to export the image, but it didn’t provide a suggested prompt. If you get a suggested prompt with this question, then select and submit it. Otherwise, manually enter this request.

  10. Select the link the agent provided to download the generated slide. Open the file once the download is complete and review the slide.