Vizuators' blog
Tableau AI Assistant
A native extension for conversational analytics
Ask your Tableau dashboards in plain English — with full transparency, native RLS, and your own LLM key
Dashboards were there, but many business questions still ended up in analysts' inboxes.
A few years ago, we started noticing the same pattern across almost every Tableau team we worked with.
Someone in the business needed a number — a simple one, like “what were sales last month?” or “which region is down?”
They’d email a BI analyst. The analyst would stop what they were doing, pull the data, and send it back.
Two days later, the business had the answer.
And the analyst had lost half a week to work that wasn’t really analysis.It wasn’t anyone’s fault. It was just the way things worked. Dashboards were there, but many business questions still ended up in analysts' inboxes.
We kept thinking:
What if people could simply ask the dashboard?
That question stayed with us. It’s what eventually led to Tableau AI Assistant.
Someone in the business needed a number — a simple one, like “what were sales last month?” or “which region is down?”
They’d email a BI analyst. The analyst would stop what they were doing, pull the data, and send it back.
Two days later, the business had the answer.
And the analyst had lost half a week to work that wasn’t really analysis.It wasn’t anyone’s fault. It was just the way things worked. Dashboards were there, but many business questions still ended up in analysts' inboxes.
We kept thinking:
What if people could simply ask the dashboard?
That question stayed with us. It’s what eventually led to Tableau AI Assistant.
What it is
Tableau AI Assistant is a dashboard extension that connects your Tableau environment to a large language model of your choice.
You ask a question. The extension figures out what data is needed, queries Tableau on your behalf, and returns an answer — all in a conversational interface that lives right inside your dashboard.
Unlike many AI tools, it was built around three principles from day one.
You ask a question. The extension figures out what data is needed, queries Tableau on your behalf, and returns an answer — all in a conversational interface that lives right inside your dashboard.
Unlike many AI tools, it was built around three principles from day one.
- TransparencyEvery answer can be verified.
With Events Mode, you can see the exact Tableau query, the returned data, the prompt sent to the LLM, and the model's response.
No black box. Just output you can inspect and validate. - SecurityThe extension uses your Tableau Personal Access Token (PAT).
Row-Level Security applies automatically, so users only see the data they are already authorized to access.
The assistant works within your existing Tableau security model rather than bypassing it. - No Vendor Lock-InYou bring your own LLM key.
Use Claude, DeepSeek, OpenAI, or the provider that meets your organization's requirements.
No vendor lock-in. No hidden AI usage costs.
Why we built it
We built Tableau AI Assistant to address three concerns we kept hearing from customers:
Earlier this year, while we were building the extension, Tableau announced its own AI agent. To us, it was a strong signal that conversational analytics is becoming a core part of the BI experience.
Our approach focuses on the dashboard as a whole, allowing users to ask questions that span multiple visualizations, filters, and metrics in a single conversation.
- AI hallucinations
- Security risks
- Vendor lock-in
Earlier this year, while we were building the extension, Tableau announced its own AI agent. To us, it was a strong signal that conversational analytics is becoming a core part of the BI experience.
Our approach focuses on the dashboard as a whole, allowing users to ask questions that span multiple visualizations, filters, and metrics in a single conversation.
We've already seen the product tested in one of the environments where trust matters most: banking.
A major bank independently evaluated the assistant, focusing on answer reliability and access controls. After completing their review, the team approved an internal rollout.
That gave us confidence that what we built wasn’t just an interesting idea — it could work in environments where security and trust are non-negotiable.
A major bank independently evaluated the assistant, focusing on answer reliability and access controls. After completing their review, the team approved an internal rollout.
That gave us confidence that what we built wasn’t just an interesting idea — it could work in environments where security and trust are non-negotiable.
What it can do
-
Natural language queries
Ask questions like:"Which regions showed a margin drop over 5% last quarter?"and get answers grounded in your Tableau data. -
Conversational follow-ups
The assistant remembers context within a session, so you can continue exploring without repeating yourself. -
Answers grounded in Tableau data
Answers are based on data retrieved from Tableau, helping users stay anchored to the metrics and business context behind each dashboard. -
Full-dashboard scope
Unlike tools limited to a single worksheet, Tableau AI Assistant works across the entire dashboard.
How to start?
Tableau AI Assistant is free to download. You bring your own LLM key — Claude, Deepseek, or others. No vendor lock-in. No hidden costs.
We built this because we wanted a tool we could actually trust. Maybe you do too.
Tableau AI Assistant is free to download. You bring your own LLM key — Claude, Deepseek, or others. No vendor lock-in. No hidden costs.
We built this because we wanted a tool we could actually trust. Maybe you do too.
Want to see how this works on your dashboards?
Let‘s walk through your setup — no commitment, just a conversation.
By clicking the button, you agree to our Privacy Policy. We‘ll use your email only to respond to your inquiry.
Contacts