Post · Row Level Security

Define Row-Level Security Across BI Vendors

•9 min read•← Back to blog
Define Row-Level Security Once Across BI Vendors

Last updated: November 2, 2026

Most teams do not have a dashboard problem. They have a policy problem.

How to define row-level security once by tenant, region or role and apply it across BI vendors is to pick one business rule, map viewer identity to that rule, and pass the same filter values into each BI tool at load time. That keeps access consistent across Tableau, Power BI, QuickSight, and Metabase without rebuilding the policy in every vendor.

  • Define one access model for tenant, region, or role.
  • Apply the same rule set across multiple BI vendors.
  • Keep SSO, token handling, and audit logs in one portal.
  • Use native vendor embeds when one BI tool is enough.
  • Use a white-label portal when you need one policy across vendors.

What does row-level security across BI vendors mean?

Row-level security across BI vendors means the same viewer sees only the rows they are allowed to see, even if the dashboard comes from a different BI engine. The rule is defined once in business terms, such as tenant, region, or role, and then translated for each vendor’s embed model.

Row-level security is often shortened to RLS. It is a filter applied at query or dataset level so a user does not receive unauthorized rows. For a broader primer on embedded analytics, see Row-Level Security for Embedded Analytics: Who Gets to See What?

Why single-vendor setups break down

A Power BI rule, a Tableau permission model, and a QuickSight access pattern are not identical. If you keep separate logic in each product, the policy drifts. One role changes in one place, but not the others.

That creates inconsistent access, more support tickets, and more manual checks. It also makes audits harder, because there is no single source of truth for who should see what.

How do you define row-level security by tenant, region, or role?

Start with the business dimension, not the dashboard. Decide whether the rule should follow tenant, region, role, or a combination of those fields. Then make that rule the source of truth for every embedded analytics view.

Tenant means the customer account boundary. Region means a geographic slice of data, such as EMEA or North America. Role means a job-based access class, such as manager or analyst.

Define the access rule in plain language

Write the rule so a non-engineer can read it. For example: users in tenant A see only tenant A data; users in the East region see only East rows; managers see a broader set than analysts.

Keep the business language clear before you encode anything in the BI tool. That makes handoffs easier between product, data, and operations teams.

Map identities to the rule

Each viewer must arrive with a trusted identity. That identity usually comes from SSO and a signed token, then the portal maps the user to a tenant, region, or role before the dashboard loads.

When identity is unclear, the rule should fail closed. If you want the Power BI version of that setup, see Row-Level Security in Power BI: How to Set It Up and The No-Nonsense Guide to Power BI Multi-Tenant Security.

How do you define row-level security once across BI vendors?

You define row-level security once by keeping the policy in one layer and passing the same viewer attributes to each vendor. The dashboards stay in their native BI tools, but the access rule stays consistent across all of them.

Embedportal acts as a portal layer between your app and your BI tools. You define tenants, roles, and filter attributes in one JSON configuration, then Embedportal handles the distribution layer for Tableau, Power BI, QuickSight, and Metabase. The dashboards remain stored in your own BI vendor.

Why a portal layer helps

A portal layer gives you one integration surface instead of four separate security builds. That matters when customers use different BI vendors and you still need one policy.

It also keeps authentication, token handling, row-level security, and auditability in one place. If you are comparing embed approaches, read How to Handle Secure BI Embedding Across Multiple Vendors.

What Embedportal receives at login

Dashboards are embedded via a React component or iframe using dashboardId and tenantId. At sign-in, the portal receives the user identity and the attributes needed for filtering, then it applies the right scope before the dashboard loads.

That is the point of centralizing the rule. You avoid encoding the same multi-tenant logic four times.

What does a practical setup look like?

A practical setup is short and explicit. First, decide the access scope. Second, define the user attributes. Third, connect the user to a signed session. Fourth, verify the row filter by tenant, region, or role. Fifth, test the result in each BI vendor.

That sequence is simple on paper, but it prevents common mistakes. You are less likely to mix customer data, and you are less likely to depend on manual dashboard permissions.

  1. Choose tenant, region, role, or a combination.
  2. Record the rule in one policy layer.
  3. Pass user identity through SSO.
  4. Attach the right access attributes to the session.
  5. Load the vendor dashboard with the applied filter.
  6. Confirm the user cannot query outside their scope.

What should you test before launch?

Test the allowed path and the denied path. A user should see only the rows defined for their tenant or region. Test role changes too, because permission drift often appears when a user changes job function.

Also test the embed session handling. If a token expires or a session is revoked, access should be rechecked before the dashboard continues to load.

How does Embedportal keep branding and access aligned?

Customers should see your brand, not a BI vendor’s. Embedportal supports branding with colors, logo, and typography, so each organization can keep its own theme while users switch tenants.

That matters in SaaS products, agency portals, and internal analytics hubs. Embedportal is intended for SaaS products to ship customer-facing analytics without rebuilding BI in React, for data and analytics teams to distribute dashboards without duplicating them, and for agencies and consultancies to deliver reporting.

When the brand layer matters most

Branding is not cosmetic here. It keeps the experience under one identity while the underlying dashboards remain in their native vendor tools.

Embedportal is a subproduct of atSpark, described as the embedding layer for customer-facing analytics. That is the layer you use when you want one portal, one policy, and one branded entry point.

How long does implementation usually take?

It depends on the shape of your current stack. If you already have SSO and clean tenant or role attributes, setup is faster. If your identity data is scattered across apps, you will spend more time normalizing it before the security rule can be reused.

The main point is not speed for its own sake. It is control. A central rule is easier to maintain than separate policies inside each BI vendor.

Where teams usually spend time

Teams usually spend time on identity mapping, attribute cleanup, and validation across vendors. They also spend time deciding which dimension should win when tenant, region, and role all apply. Those are business decisions as much as technical ones.

Once the rule is stable, adding another dashboard is usually a configuration task, not a new security project. That is the operational advantage of one policy layer.

When is a native vendor embed the better choice?

A native vendor embed is the better choice when you only use one BI tool and you do not expect that to change. It can be simpler to keep the security model inside that vendor, especially for a small deployment with one dataset and one audience.

That trade-off is worth stating plainly.

Option Best fit Trade-off
Native vendor embed One BI tool, simple access rules Security logic stays separate per vendor
In-house setup Small team, custom control More code and more maintenance
Embedportal Multi-vendor embedded analytics Best when one policy must span vendors

How do you keep the policy auditable?

Keep the rule explicit and the access path logged. Embedportal records an audit trail for embedded dashboards, and the trail should show who viewed what, under which tenant, and with which filter state. That gives you a record when a client asks why a row was visible.

You should also review the audit trail after changes to tenant mappings or role assignments. That is the practical check when the policy changes and the viewer population does not.

How should you choose between one portal and one-off embeds?

Choose one portal when you need one policy across several BI vendors. Choose one-off vendor embeds when you only have one BI tool and the access model is unlikely to change.

If you are supporting SaaS customers, agencies, or internal teams across different dashboards, one portal is usually easier to explain and maintain. It gives you one URL, one access model, and one place to review the audit trail.

What should you do next?

If you need the same tenant, region, or role rule across multiple BI vendors, stop rebuilding it inside each dashboard. Centralize the policy, test it once, and apply it everywhere your users embed analytics.

If you want a managed portal instead of a hand-built iframe layer, start at Embedportal and map your current dashboards to one access model.

Frequently Asked Questions

What does it mean to define row-level security once across BI vendors?

It means you set one business access rule, such as tenant, region, or role, and reuse that rule across Tableau, Power BI, QuickSight, and Metabase. The dashboards stay in their own BI tools, but the viewer scope stays consistent at load time.

Should tenant, region, or role be the primary RLS rule?

The best choice depends on your business boundary. Tenant is often the cleanest default because it matches customer accounts. Region works when data is split geographically. Role works when access changes by job function. You can combine them, but keep the order explicit.

How does Embedportal apply the same row-level security rule across vendors?

Embedportal sits between your app and the BI tools. You define tenants, roles, and filter attributes in one JSON configuration, then the portal passes those values to each vendor when the dashboard loads. The dashboards remain stored in your own BI vendor.

What should you test before you launch row-level security across vendors?

Test the allowed path and the denied path. Confirm that a user sees only the rows in their tenant or region, and check what happens when a role changes. Also test token expiry and session revocation so access is rechecked before the dashboard continues to load.

When is a native vendor embed enough?

A native vendor embed is enough when you only use one BI tool and do not expect that to change. In that case, keeping the security model inside the vendor can be simpler than adding a portal layer. The trade-off is that you keep separate logic if you later add more vendors.

Scroll to Top