Support

Tableau Server 2025.1+ or Tableau Cloud.

No. Any user with edit permissions on a dashboard can add the extension.

Yes. The extension works with any dashboard that has a published data source accessible via the user‘s PAT. It does not require modification to the dashboard itself.

Currently Anthropic Claude and Deepseek. Both work well; the choice depends on your existing relationships, pricing preferences, and any internal AI vendor policies your organization may have. More providers will be added regularly.

Yes. The extension stores credentials locally in your browser. You can switch providers at any time by updating the API key in the extension settings.

The extension handles aggregations, filters, comparisons, top-N queries, conditions, and time-series trends well. Examples that work reliably:

• “What is total revenue for Q3?”
• “Top 10 customers by ARR in the Tech industry”
• “Compare margin this quarter vs last quarter by region”
• “Which products had sales decline of more than 5% last month?”

No. We never use any user data to train any models, full stop.

Natively. The extension uses your Tableau PAT, so RLS defined in Tableau applies directly. Users never see data they're not authorized to access.

Locally in your browser's local storage. They are never stored on our servers.

Yes. Toggle Events mode in the extension interface. This reveals the query sent to Tableau, the data returned, the prompt sent to the LLM, and the model‘s raw response.

No. Chat history is session-only. When you close the dashboard or refresh the page, the conversation context is cleared. Nothing is stored on our infrastructure.

Yes. Each user installs the extension and configures it with their own Personal Access Token and LLM API credentials. Each user‘s queries are executed under their own permissions.

Yes, there is a Free tier. Pro tier with additional features is coming soon — join the waitlist on our Pricing page.

Yes — those are your accounts. We don't mark up or resell LLM access.

Pro is currently in waitlist. Early-bird pricing will be available for the first 100 teams. Join the waitlist.

Each user provides their own PAT. RLS applies per user — automatically. No shared credentials.

Enable Events mode (toggle in the extension). You'll see the exact query sent to Tableau, the data returned, and the prompt to the LLM. This helps debug and verify every step.

No. Tableau Server 2025.1+ or Tableau Cloud is required.

Verify three things:

1. Tableau version: 2025.1+ or Tableau Cloud. Check via Help → About Tableau.
2. HTTPS: the extension is hosted over HTTPS. If your Tableau Server blocks mixed content, ask your admin to whitelist our host.
3. Browser: modern versions of Chrome, Edge, Firefox, Safari. Internet Explorer is not supported.

Most common causes:

• PAT expired (default lifetime: 180 days)
• PAT name or value entered incorrectly (case-sensitive)
• PAT generated for a different site than the dashboard‘s site
• User account lacks data source access

Verify:

• API key is valid (test in the provider‘s playground)
• Account has credit / active billing
• Rate limits — wait a few minutes or upgrade your LLM plan

Enable Events mode to inspect what happened:

1. Check the query — did the extension build a query matching your intent?
2. Check the data returned — did Tableau return what you expected?
3. Check the prompt to the LLM — did the LLM receive enough context?
4. Check the LLM response — did the model misinterpret?

This is almost always a Row-Level Security restriction. The extension queries Tableau on your behalf using your PAT — the data you see is exactly the data you would see if you queried Tableau directly. Check with your Tableau administrator.

Response time depends on:

• Tableau query speed (your Tableau Server / data source performance)
• LLM response time (typically 2–5 seconds)
• Network latency (we aim to keep overhead under 200ms)

The extension routes requests through three parties:

1. Your Tableau Server / Tableau Cloud — queried using your PAT, with all access controls applied natively.
2. Our processing infrastructure — translates between Tableau and the LLM provider. Data is processed in-memory only; nothing is written to persistent storage.
3. Your chosen LLM provider — receives your question, query results, and dashboard context.

Depends on your question. Typically:

• Your natural-language question
• Metadata about the dashboard‘s data source (field names, formulas, structure)
• Query results from Tableau relevant to answering your question

Revoke the PAT in your Tableau account settings. This invalidates all sessions using that PAT. Generate a new PAT and re-enter it in the extension.

Not in the current version. The extension requires internet access. For organizations requiring fully on-premises AI deployment, contact us at info@vizuators.com.

We are working toward formal certification. In the meantime, we provide a documented data flow, defined architecture with no persistent storage of user data, TLS 1.2+ encryption, and a security disclosure channel.

Yes — early access will be available for waitlist members.

Persistent chat history, shared prompts within your team, team workspace and admin controls, priority support with SLA, additional LLM providers (GPT-4o, Gemini, Llama 3), and early access to new features.