Setup

Prerequisites

Before connecting, ensure the following are in place:

The workspaces Prizm will scan are hosted on Power BI Premium Per User (PPU), Premium capacity, or Microsoft Fabric capacity. Prizm reads semantic model metadata (tables, columns, measures, relationships, calculated fields) over the XMLA endpoint, which is only available on these capacity tiers — a standard Pro license alone is not sufficient.
The XMLA endpoint is enabled with Read access for metadata discovery (Power BI Admin portal → Tenant settings → Integration settings). Write access is not required.
The Prizm service identity has been granted the required API permissions in Azure Active Directory (see below).
Admin API access is enabled in the Power BI tenant settings: go to the Power BI admin portal → Tenant settings → Admin API settings and enable Allow service principals to use read-only Power BI admin APIs.
Fabric API access is enabled for service principals: in the Fabric Admin portal, go to Tenant settings → Developer settings and enable Service principals can call Fabric public APIs.
If you plan to use OAuth 2.0 authentication, a redirect URI is registered on the Azure AD app registration: https://<your-prizm-domain>/source/powerbi/callback.

Required API Permissions

Prizm accesses Power BI through the Power BI REST API and the Power BI Scanner API. The following delegated or application permissions are required in Azure AD, depending on your authentication method:

Permission	Purpose
`Dataset.Read.All`	Read semantic model metadata, tables, columns, measures
`Dashboard.Read.All`	Read dashboard tiles and metadata
`Report.Read.All`	Read report metadata and page structure
`Workspace.Read.All`	List and enumerate all workspaces
`Dataflow.Read.All`	Read dataflow metadata and M-expressions

All permissions above are read-only. Prizm does not write to, modify, or trigger refreshes on your Power BI assets during catalog or observability runs.

Connector Setup

Open the connector wizard

Navigate to Settings → Connectors and click Add. Select Power BI from the source type grid.

Field	Description
Connection Name	A label for this connection (e.g., `powerbi-prod`)
Description	Optional notes about what this connection covers

Choose an authentication method

Select your authentication method from the dropdown. The three supported methods are described below.

Service Principal (Recommended)
Master User
OAuth 2.0

Service Principal is the recommended authentication method for production deployments. It authenticates as a registered Azure AD application rather than a named user, which means:

Credentials are not tied to an individual’s account or license
Permissions survive employee offboarding or license changes
Audit logs show a distinct service identity rather than a user account

Azure-side setup:

Go to Azure Portal → Azure Active Directory → App registrations → New registration. Give it a name (e.g., prizm-powerbi-connector).
Under Certificates & secrets, create a Client secret. Copy the value immediately — it is shown only once.
Under API permissions, add the five permissions listed in Prerequisites as Application permissions and grant admin consent.
Note the Application (client) ID and Directory (tenant) ID from the app’s overview page.
In the Power BI Admin portal → Tenant settings → Admin API settings, add the service principal’s security group (or the app itself) to the allowed list.

Then enter the following in Prizm:

Field	Description
Tenant ID	Azure AD directory (tenant) ID
Client ID	Application (client) ID of the registered app
Client Secret	The secret value generated in step 2

Workspace access: In addition to the Azure AD permissions above, the service principal must be added directly to each target Power BI workspace (workspace Manage access → add the app/service principal) with the Member or Admin role. It also needs Build permission on the semantic models Prizm will scan — this is required for XMLA metadata access.

The same Entra ID app registration can be reused across Service Principal, Master User, or OAuth authentication, depending on which method you select for this connection.

Master User authenticates with a standard Power BI user account (username and password). The account must have a Power BI Pro or Premium Per User license and must be a member of all workspaces you want Prizm to access.

Master User authentication is not recommended for production. If the user’s account is disabled, their password changes, or MFA enforcement changes, the connector will fail silently until credentials are updated. Use Service Principal for stable, long-running connections.

Field	Description
Username	Email address of the Power BI user account
Password	Account password

Requirements:

The account must not have MFA enforced (or must be excluded from MFA conditional access policies for service-account use).
The account must have at least Power BI Pro license.
Add the account as a Member or Admin in each workspace you want Prizm to catalog.

OAuth 2.0 uses the delegated authorization flow to connect Prizm as the authenticated user who grants access. This is suited for single-user or evaluation setups where a Service Principal is not yet available.Before authorizing, make sure the Azure AD app registration has the following redirect URI configured: https://<your-prizm-domain>/source/powerbi/callback.Click Authorize with Microsoft in the Prizm connector form. You will be redirected to Microsoft’s login page to grant the required permissions. Once you approve, Prizm stores the refresh token and renews access automatically.Limitations:

Access is bound to the authorizing user’s account and workspace memberships.
If the user’s account changes (password reset, license change, MFA), the token may be invalidated.
Not recommended for shared or production environments — use Service Principal instead.

For production deployments, always use Service Principal. It avoids credential decay, works with tenant-level admin API access, and produces a clean audit trail.

Configure extraction scope

After authentication is verified, select which workspaces to include:

Setting	Description
Workspaces	Select specific workspaces by name, or leave blank to include all workspaces the service principal has access to
Extract Dataflows	Toggle on to include Power Query dataflows in the catalog and lineage graph. Disabled by default.

Enabling Extract Dataflows increases catalog job duration proportionally to the number of dataflows in scope. Enable it only for workspaces where dataflows are part of the lineage you want to trace.

Review job schedules

For Power BI connections with semantic models in scope, Prizm creates the following background jobs automatically. You can adjust their schedules from the connector detail page after saving:

Job	Default Cadence	What it collects
Technical	Daily — first job in the pipeline	Workspace metadata and semantic model schema (tables, columns, measures, calculated columns, relationships, hierarchies) via the XMLA endpoint, plus reports, dashboards, dataflows, and data sources via the REST/Scanner API
Operational	Hourly	Volume, freshness, and schema-change signals on semantic models, including refresh history and success/failure status
Performance	Daily	Refresh duration trends and capacity/query performance signals for semantic models
Profile Recommendation	Runs in parallel after Operational completes	Determines which semantic model columns to profile, at what frequency, based on criticality and scope. The Profile job will not run until this job produces its output.
Profile	Runs on the schedule generated by Profile Recommendation	Column-level statistical profiling on semantic model columns — null rate, distinct count, min/max, mean, pattern conformance, and related metrics
Structural Recommendation	Runs in parallel after Operational completes	Recommends structural quality metrics (Essential or Advanced) for each semantic model column, based on data type and criticality. The Structural job executes against this output.
Structural	Triggered after Structural Recommendation completes	Distribution, frequency, pattern, and statistics metrics at the column level for in-scope semantic model columns

This job set applies to Power BI connections where semantic models are in scope. Governance, Usage Summary, Compute Metric, and Lineage jobs are not part of the current Power BI job pipeline. Operational, Technical, and Performance job execution has been verified directly against a live Power BI connector.

Manual re-sync: Trigger a manual catalog run at any time from Settings → Connectors → [your connection] → Run Now. Manual runs execute immediately without disrupting the configured schedule.

Save and verify

Click Save to create the connection. Prizm runs an initial full catalog scan immediately. Assets appear in the catalog within a few minutes depending on workspace size.Navigate to Settings → Connectors → Logs to monitor scan progress and review any errors from the initial run.

Next Steps

What We Collect

See the full field-level breakdown of every metadata object Prizm extracts from Power BI.

FAQ

Common questions about authentication, scoping, and permissions.

Get Started

Core Concepts

Platform

Architecture

Data Sources

Integrations

Security & Access Control

Deployment

Help

Settings

MCP

Release Notes

API Reference

Prerequisites

Required API Permissions

Connector Setup

Next Steps

What We Collect

FAQ

​Prerequisites

​Required API Permissions

​Connector Setup

​Next Steps

What We Collect

FAQ

Prerequisites

Required API Permissions

Connector Setup

Next Steps