Add-on · Power BI Embedded

WFS PowerAutoScale — AI picks the SKU of your Power BI Embedded capacity

Painel WFS PowerAutoScale — capacidade Power BI Embedded em A2 com auto-escala

Smart auto-scaling for Power BI Embedded (SKUs A1–A8) with predictive AI, dry-run mode and dynamic ranges. Your capacity stays in your Azure tenant — WFS just manages it. Ideal for studies, POC, staging and real production.

Predictive AI (+5 min) DRY-RUN mode Dynamic ranges Your capacity, your tenant
The problem

Power BI Embedded is billed by the hour — at fixed SKU, you pay peak rates 24/7

Embedded capacities (A1–A8) cost per uninterrupted hour. Fixing at A4 to handle the peak means paying A4 at night, weekends and dawn too — when nobody is using it. Result: spend that could be 3–10× lower.

How it works

You bring the capacity, AI handles scaling

1
Register your capacity Add the Subscription/Resource Group and grant WFS access via Azure RBAC. Step-by-step manual included.
2
AI watches CPU in real time Polls every ~1 min via Azure API. Applies outlier filtering and predicts +5 min ahead.
3
Decides the ideal SKU Compares predicted CPU against dynamic ranges (configurable) and picks the A1–A8 that fits the expected peak.
4
Executes scale on Azure Calls Microsoft's ARM API to scale/pause. DRY-RUN mode simulates everything without applying — great for validation.
Important — your capacity, your tenant. The Power BI Embedded capacity is provisioned in your Azure account and Azure costs (SKU × hour) are billed directly by Microsoft. WFS PowerAutoScale only decides which SKU to run — does not resell capacity nor add markup to Azure. You pay the WFS license per managed capacity; Azure cost stays transparent on your Microsoft invoice.
A1 1 v-core · ~3 GB
A2 2 v-cores · ~5 GB
A3 4 v-cores · ~10 GB
A4 8 v-cores · ~25 GB
A5 16 v-cores · ~50 GB
A6 32 v-cores · ~100 GB
A7 64 v-cores · ~200 GB
A8 128 v-cores · ~400 GB
Who it's for

Designed for scenarios where usage varies throughout the day

📚 Studies and training

Students, instructors, Microsoft certs (PL-300, DA-100). Up only during class, A1 in breaks, off at night. Studying Power BI Embedded becomes viable without burning A4 24/7 budget.

🧪 POC and staging

Technical validation before committing to a fixed monthly SKU. Executive demo, client sandbox, staging that only runs in short test windows.

🏢 Real production

Corporate tenants with predictable peaks (business hours, month-end). Britânia case: A4 at peak, A1 off-peak — over 60% monthly savings with no performance loss at critical hours.

Azure setup

~15-minute setup with step-by-step manual

The capacity lives in your Azure tenant. You grant minimum RBAC (Contributor scoped to the capacity's Resource Group) to the WFS Service Principal, register Subscription ID + Resource Group + capacity name in the dashboard, and the scaler starts in DRY-RUN. When you're confident, flip the switch to live in 1 click.

1. Service Principal

Create an App Registration in your Azure AD or use the SP provided by WFS. Permission: Microsoft.PowerBIDedicated/capacities/* scoped to the Resource Group.

2. Register in the dashboard

In the WFS PowerAutoScale dashboard, register: Subscription ID, Resource Group, capacity name, ranges (min/max SKU), usage windows.

3. DRY-RUN → production

For ~24h the system simulates decisions without applying. You check projected savings vs baseline, tune ranges, and flip real scaling whenever you want.

Full technical manual with screenshots, ready-to-use az cli commands and permission troubleshooting delivered after subscription. Email support during setup.

License

Simple billing: per managed capacity

No complicated tiers. No trial. Register one or many capacities — pay per each. Add or remove anytime via Stripe Customer Portal.

WFS PowerAutoScale license

R$ 1.630/capacity/month

or US$ 329/capacity/month outside Brazil · billed monthly

  • AI picks SKU (A1–A8) with +5-min prediction
  • DRY-RUN mode for validation before applying
  • Dynamic ranges per capacity and time window
  • Dashboard with monthly savings vs baseline (fixed SKU)
  • History of AI decisions with rationale
  • Azure setup manual + email support
  • Stripe Customer Portal: add/remove capacities yourself — automatic prorated billing
  • No lock-in · cancel anytime
  • Annual commitment · cancel for next cycle
Contact us to activate

Azure cost (SKU × hour) billed directly by Microsoft on your Azure account — does not go through WFS, no markup.

How it works when you register a new capacity

  1. In the WFS dashboard you click "Add capacity" and enter Subscription ID + Resource Group + capacity name.
  2. The dashboard shows a cost preview: "Your subscription has 2 capacities — adding this makes it 3, totaling 3 × R$ 1,630 = R$ 4,890/month".
  3. You confirm. The backend calls Stripe and increments the subscription quantity; Stripe immediately debits the prorated difference on the saved card.
  4. Once payment is approved, the new capacity is enabled automatically and AI starts observing in DRY-RUN.
  5. To remove a capacity, the reverse flow: click remove → confirm → quantity decremented on Stripe → prorated credit on next cycle.

Multiple capacities, groups or holdings?

For 5+ capacities, cross-tenant balancing, dedicated SLA or multi-capacity reinforced AI, we design a custom contract.

Talk to sales
FAQ

About WFS PowerAutoScale

Is the Power BI Embedded capacity WFS's or mine?

Yours. The capacity is provisioned in your Azure tenant, under your Subscription, and Azure cost (SKU × hour) hits your Microsoft invoice directly — no markup. WFS PowerAutoScale only decides which SKU to run via ARM API. You own the capacity end-to-end; you can revoke the WFS SP permission anytime.

How does WFS license billing work?

R$ 1,630 per managed capacity per month (or US$ 329 outside Brazil), via Stripe — card, Pix or boleto. The subscription has an adjustable quantity: 3 capacities = 3 × R$ 1,630. Choose monthly (default) or annual (~10% off). Adding/removing capacities is done in the WFS dashboard — Stripe bills the prorated difference automatically on the card.

When I register a new capacity, how am I billed?

The dashboard shows a preview before confirmation: "You have 2 capacities. Adding this brings your next bill to 3 × R$ 1,630 = R$ 4,890/month". On confirm, the backend bumps your Stripe subscription quantity; Stripe immediately debits the difference prorated to the remaining cycle on the saved card. Once payment clears (seconds), the new capacity is enabled and enters DRY-RUN. Removal is the reverse — prorated credit lands next cycle.

Monthly or annual — which to pick?

Monthly is the default (R$ 1,630/cap/month), no lock-in. Annual equals R$ 1,467/cap/month (10% off) — you pay R$ 17,604 per capacity upfront covering 12 months, saving R$ 1,956/cap/year. Continuous use (production) pays back fast. For short studies, monthly gives more flexibility. You can migrate monthly→annual anytime via Customer Portal.

Is there a free trial?

No. Since the product starts in DRY-RUN (simulates decisions without applying), you validate the scaler's value in the first 24–48h after subscribing — and only flip real scaling when comfortable. Cancel in the first 7 days and we refund the full monthly fee.

Does it make sense for studying Power BI Embedded?

Very much. Studying/certifying (PL-300, DA-100, DP-500) requires real Embedded to test Premium features. An A4 24/7 costs thousands of dollars per month — unfeasible for students. With PowerAutoScale, run A1–A4 only during class/lab, A1 at idle, pause outside. Azure cost drops sharply; you only pay R$ 1,630/month WFS license.

Can the capacity be paused?

Yes. AI can pause (capacity Paused state) in zero-usage windows — Azure doesn't bill while paused. You define in the dashboard which windows are safe to pause (e.g. dawn, weekends).

What if AI picks wrong and my dashboard hangs?

That's why DRY-RUN exists. Before applying real scaling, you run 24–48h simulating and check if AI would react well to peaks. Dynamic ranges (min/max SKU) are configurable per capacity — force minimum at A2 if A1 is too low. Full decision history with rationale is in the dashboard.

Do I need to already have Power BI Premium / Embedded?

Yes. PowerAutoScale is an add-on to Power BI Embedded — it does not replace it. You need an Azure Subscription with at least one Embedded capacity (A1+) provisioned. If you don't have it yet, WFS can help via consulting.

Works with Microsoft Fabric (F-SKU) or only Embedded (A-SKU)?

Today, official support is A-SKU (A1–A8). Fabric F-SKU has a different API and scaling model; on the roadmap but not GA yet. Talk to sales if Fabric is a hard requirement now.

Start in DRY-RUN today

Subscribe, set up in ~15 min with our manual, and within 24h see AI's projected savings on your real capacity.

Subscribe now — R$ 1,630/capacity