A catalogue of agents in your tenant from the Agent Registry in the Microsoft 365 Admin Center. This input is optional — the template loads without it.

If you ran PAX with -IncludeAgent365Info, PAX already exported this file. Skip the manual steps below — PAX output goes directly into the template parameter.

Access required: AI Admin or Global Reader

1. Go to admin.microsoft.com → Agents → All Agents
2. Click Export to Excel in the toolbar
3. Save the .xlsx file to a known location (e.g., C:\Data\Agent365_Inventory.xlsx)

• File format: Excel (.xlsx)
• Columns: Name, Host products, Created date, Developer user ID, Description, Status, Version
• Rows: One row per agent in your tenant

Open AI-In-One - Rollup Edition.pbit in Power BI Desktop.

Each parameter accepts a local file path, SharePoint URL, or OneLake URL — the template auto-detects the source type.

Click Load. First refresh on a moderate dataset typically takes 5–15 minutes.

See the Rollup Edition README for SharePoint URL conventions, OneLake path format, scheduled refresh setup, and full troubleshooting.

• "File not found" / DataSource.Error — local paths must be absolute (e.g. C:\Data\file.csv, not .\file.csv). For SharePoint URLs, copy the document URL, not the share link.
• Formula.Firewall: Query references other queries… — privacy-level mismatch. In Power BI Desktop: File → Options → Current File → Privacy → Combine data without privacy. In Service: dataset Settings → Data source credentials → set Privacy: None for SharePoint sources.

Review the dashboard, customize visualizations, and share with stakeholders.

1. Review dashboard pages

   • Navigate through all report pages
   • Verify data loaded correctly
   • Check that filters and slicers work as expected

2. Customize for your organization

   • Update visuals to match your branding (colors, logos)
   • Adjust hierarchies to match your org structure
   • Add or remove pages based on your needs
   • Create bookmarks for common views

3. Set up filters and parameters

   • Configure default date ranges
   • Set up department/role filters
   • Create user-specific views if needed

4. Publish and share

   • Publish to Power BI Service if not already done
   • Set up Row-Level Security (RLS) if needed
   • Share with stakeholders via workspace access or apps
   • Create subscriptions for key reports

5. Document customizations

   • Keep notes on any changes you make
   • Version your .pbix file if making significant updates
   • Archive old versions in the /Archived Templates folder

• 🔄 Refresh schedule: Schedule PAX (or your standalone processor run) and the Power BI dataset refresh on the same cadence so the report always reflects current data
• 🔒 Security: Use Row-Level Security to restrict sensitive data by department or role
• 📧 Subscriptions: Set up email subscriptions for executives who want regular updates
• 📊 Usage tracking: Monitor dashboard usage in Power BI Service to understand what resonates

Follow the traditional workflow using browser-based portals to export your data:

1. Export Copilot audit logs from Microsoft Purview
2. Download licensed user data from Microsoft 365 Admin Center
3. Export org data from Microsoft Entra Admin Center
4. Connect CSV files to Power BI template

Best for: One-time setup, first-time users, or those who prefer GUI-based workflows

Use the PowerShell automation scripts in the scripts folder for a faster, repeatable workflow. This method supports two execution modes:

• Run locally (PowerShell) and export CSVs
• Run in Azure Automation (runbooks) and upload outputs to SharePoint

Advantages:

• ✅ Automated data export via Microsoft Graph API
• ✅ Reduced manual steps and potential errors
• ✅ Easy to schedule for regular data refreshes
• ✅ Consistent results every time

Prerequisites:

• PowerShell 5.1 or later
• Microsoft Graph PowerShell modules
• Appropriate permissions (same as manual method)

Quick Start (Local execution):

# 1. Install required modules
Install-Module Microsoft.Graph.Beta.Security -Scope CurrentUser

# 2. Navigate to scripts folder and run
cd scripts
.\create-query.ps1              # Creates audit log query
.\get-copilot-interactions.ps1  # Exports query results
.\get-copilot-users.ps1         # Exports licensed users list

Quick Start (Azure Automation execution):

cd scripts/automation
.\deploy.ps1

📖 Documentation:

• Local scripts: scripts/readme.md
• Azure Automation runbooks: scripts/automation/README.md

This log provides detailed records of Copilot interactions across all surfaces (Chat, M365 apps, Agents), as well as interactions with third-party and custom-built AI applications (e.g., Confluence Cloud, Jira Cloud, Miro), enabling deep analysis of usage patterns and engagement across the full AI landscape.

• Access level required: Audit Reader or Compliance Administrator
• Portal: Microsoft Purview Compliance Portal
• Permissions needed: View and export audit logs

1. Navigate to the portal

   • Go to: security.microsoft.com
   • In the left pane, scroll down and click Audit
   • Ensure you have appropriate compliance roles (e.g., Audit Reader). If not, contact your IT admin

2. Configure the audit search

   • In Activities > Friendly Names, select:

     • Copilot Activities – Interacted with Copilot (required) — M365 Copilot interactions across all surfaces (RecordType: CopilotInteraction)

   • Recommended: Also select these two additional activities to capture third-party and custom AI app usage:

     • Copilot Activities – Interacted with a Connected AI App — Custom-built Copilots and registered 3rd-party AI apps that your org has deployed (RecordType: ConnectedAIAppInteraction)
     • Copilot Activities – Interacted with an AI App — Non-Microsoft 3rd-party AI apps accessed via Microsoft 365, even if not formally deployed in your org (RecordType: AIAppInteraction)

   • Set a Date Range (recommended: 1–3 months to match your Viva query)

   • Give your search a descriptive name (e.g., "Copilot Audit Export - Oct 2025")

   💡 Why include the extra activities?
   The standard Interacted with Copilot activity only captures M365 Copilot usage. As organisations adopt third-party agents and custom Copilots (e.g., Confluence Cloud, Jira Cloud, Miro), these interactions are logged under separate record types. Including them gives you a complete picture of AI adoption — not just Microsoft Copilot, but the full ecosystem of AI tools your users are engaging with.

   ⚠️ Pay-as-you-go billing for "Interacted with an AI App" (AIAppInteraction / RT405):
   This third activity type uses Microsoft Purview pay-as-you-go (PAYG) billing and is not included in standard Audit (Standard) or Audit (Premium) subscriptions. To enable it:

   1. An admin must set up pay-as-you-go billing in Microsoft Purview, which bills based on the volume of audit records generated
   2. Once enabled, Purview begins logging interactions with non-Microsoft AI applications
   3. Costs are consumption-based — you only pay for the records actually generated

   If PAYG is not enabled, selecting this activity will simply return no results — the first two activities will still work normally. You can always add this later without re-exporting the other data.

   The second activity (Interacted with a Connected AI App) does not require PAYG — it is included with your existing Audit subscription and covers custom-built Copilots and registered 3P apps.

3. Run and export the search

   • Click Search
   • Wait until the status changes to Completed
   • Click into the completed search
   • Select Export > Download all results
   • Save the CSV file to a known location (e.g., C:\Data\Copilot_Audit_Logs.csv)

• File format: CSV
• Typical size: Varies widely (5 MB–500 MB depending on org size and activity)
• Columns: ~50+ columns including timestamps, user IDs, activity types, surfaces
• Rows: One row per Copilot interaction

📖 Learn more: Export, configure, and view audit log records – Microsoft Learn

This data provides a list of users with Copilot licenses, enabling you to track license utilization and identify licensed vs. unlicensed usage patterns.

• Access level required: Global Administrator or Reports Reader
• Portal: Microsoft 365 Admin Center
• Permissions needed: View usage reports

1. Navigate to the portal

   • Go to: admin.microsoft.com
   • Log in as a Microsoft 365 Global Administrator or Reports Reader

2. Unhide usernames (if concealed)

   • Go to Settings > Org Settings
   • Under the Services tab, choose Reports
   • Deselect: "Display concealed user, group, site names in all reports"
   • Click Save changes

3. Navigate to Copilot reports

   • Go to: Reports > Usage > Microsoft 365 Copilot
   • Click on the Readiness tab

4. Export license data

   • Scroll to Copilot Readiness Details section
   • Ensure the column Has Copilot license assigned is visible
   • Click the ellipsis (...) menu
   • Choose Export to download the file as CSV
   • Save to a known location (e.g., C:\Data\Copilot_Licensed_Users.csv)

• File format: CSV
• Typical size: 1–10 MB for 1,000–10,000 users
• Columns: ~10–15 columns including UserPrincipalName, Department, LicenseStatus, LastActivityDate
• Rows: One row per user in your organization

📖 Learn more: Microsoft 365 Copilot Readiness Report – Microsoft Learn

A catalogue of agents in your tenant from the Agent Registry in the Microsoft 365 Admin Center.

• Access level required: AI Admin (or Global Reader for view-only access)
• Portal: Microsoft 365 Admin Center

1. Go to admin.microsoft.com → Agents → All Agents
2. Click Export to Excel in the toolbar
3. Save the .xlsx file to a known location (e.g., C:\Data\Agent365_Inventory.xlsx)

• File format: Excel (.xlsx)
• Columns: Name, Host products, Created date, Developer user ID, Description, Status, Version
• Rows: One row per agent in your tenant

This file provides organizational hierarchy and user attributes, enabling segmentation by department, role, location, or other organizational dimensions.

• Access level required: User Administrator or Global Reader (Entra) OR Insights Administrator (Viva)
• Portal: Microsoft Entra Admin Center or Viva Insights
• Permissions needed: View and export user data

1. Navigate to the portal

   • Sign in to: entra.microsoft.com
   • In the left-hand navigation, go to: Identity ➝ Users

2. Select and download users

   • Click All users
   • Click the "Download users" button (in toolbar or under ... menu)

3. Configure the export

   • In the download dialog, select attributes to include:
   • Required fields:
     • UserPrincipalName
     • Department
   • Optional but recommended fields:
     • JobTitle
     • Office
     • City
     • Country
     • Manager
     • Any custom attributes relevant for reporting

4. Download the file

   • Choose CSV format
   • Click Download
   • Save to a known location (e.g., C:\Data\Org_Data_Entra.csv)

If you have a custom org data file with organizational hierarchy and user attributes, you can use that instead. Ensure it includes:

• Required columns: UserPrincipalName or PersonID, Department or Organization

• File format: CSV
• Typical size: 1–20 MB depending on org size and attributes
• Columns: Varies (5–30+ columns)
• Required columns: UserPrincipalName, Department
• Rows: One row per user

💡 Note: Avoid downloading non-essential attributes as it can degrade performance and increase file size unnecessarily.

📖 Learn more: Download a list of users – Microsoft Learn

Pick the deployment path that matches where your CSVs live, open the matching PBIT, and supply parameters.

Refer back to the Choose your deployment path table above. The available templates:

Each per-folder README has the definitive setup steps, parameter values, and troubleshooting for that variant — including SharePoint folder URL conventions, Service refresh / credential setup, and (for Fabric) the upstream notebook + Lakehouse provisioning.

Rollup Edition — 3 parameters:

Each parameter accepts a local file path, SharePoint URL, or OneLake URL — the template auto-detects the source type.

See the Rollup Edition README for full details on both PAX and standalone processor paths, SharePoint URL guidance, and scheduled refresh setup.

Classic Edition — 4 parameters:

Only the value format changes between Classic Edition paths (local path vs SharePoint URL vs SharePoint folder vs Lakehouse table). See each path's README for the parameter values specific to that variant.

After supplying parameters, click Load. First refresh on a moderate dataset takes 5–15 minutes; large tenants on the Manual / SharePoint paths may need 30+ minutes (use the Fabric path if you're hitting the 1 GB / 2-hour Service caps).

• "File not found" / DataSource.Error — local paths must be absolute (e.g. C:\Data\file.csv, not .\file.csv). For SharePoint URLs, copy the document URL, not the share link.
• Refresh times out in Power BI Service (>2 hours) — switch to the 3. Fabric/ deployment path. JSON parsing happens upstream in a Lakehouse / notebook, leaving the dataset as a thin pass-through against a flat Delta table.
• Formula.Firewall: Query references other queries… — privacy-level mismatch when combining sources. In Power BI Desktop: File → Options → Current File → Privacy → Combine data without privacy. In Service: dataset Settings → Data source credentials → set Privacy: None for SharePoint sources.
• Path-specific issues — see the troubleshooting section in each per-folder README.

Review the dashboard, customize visualizations, and share with stakeholders.

1. Review dashboard pages

   • Navigate through all report pages
   • Verify data loaded correctly
   • Check that filters and slicers work as expected

2. Customize for your organization

   • Update visuals to match your branding (colors, logos)
   • Adjust hierarchies to match your org structure
   • Add or remove pages based on your needs
   • Create bookmarks for common views

3. Set up filters and parameters

   • Configure default date ranges
   • Set up department/role filters
   • Create user-specific views if needed

4. Publish and share

   • Publish to Power BI Service if not already done
   • Set up Row-Level Security (RLS) if needed
   • Share with stakeholders via workspace access or apps
   • Create subscriptions for key reports

5. Document customizations

   • Keep notes on any changes you make
   • Version your .pbix file if making significant updates
   • Archive old versions in the /Archived Templates folder

• 🔄 Refresh schedule: Set up weekly or monthly refresh for CSV files in Power BI Service
• 🔒 Security: Use Row-Level Security to restrict sensitive data by department or role
• 📧 Subscriptions: Set up email subscriptions for executives who want regular updates
• 📊 Usage tracking: Monitor dashboard usage in Power BI Service to understand what resonates