VantegrateNative API integrationIntegrate Tableau with AI agentsthat answer questions from your governed data
Vantegrate's AI agents connect to your site through the official Tableau APIs (REST and the VizQL Data Service): they query your published data sources, answer what a dashboard shows and explain what changed, in natural language. Your governed data keeps living in your Tableau Cloud or Tableau Server.
How the data flows
What does it mean to integrate Tableau with Vantegrate's AI agents?
Integrating Tableau with Vantegrate's AI agents means that a data agent (Metrix) connects to your site through the official Tableau APIs and queries the governed data in your published data sources. You ask in natural language, the agent turns that question into a VizQL Data Service query and gives you the figure with its context, without opening Tableau.
That is different from exporting a report to a spreadsheet that goes stale. Here the agent reuses the same governed metrics that feed your dashboards (your data team's source of truth): it queries the VizQL Data Service for the number and the Metadata API for the lineage. The data stays in your Tableau; the agent queries it and explains it, it doesn't take it anywhere else.
What do you want to connect to your Tableau?
Pick what you need and we'll write the message for you.
1 · Pick what to solve
2 · What our team receives
Vantegrate
online
What data the agent reads in your Tableau
The agent works on your published data sources at query time: it reads the governed metric to answer a question, explains what changed and resolves the lineage, without rebuilding your models.
Published data sources
The agent queries the data in a published data source through the VizQL Data Service, with a JSON request that filters and aggregates, without anyone having to build a new visualization.
What changed in your views
The agent reads the data source periodically and explains the deviation, the spike or the drop in natural language, so the team learns about the change without checking every dashboard by hand.
Data catalog and lineage
With the Metadata API, the agent answers where a number comes from and which dashboards use a table, so governance and traceability are available over chat.
How the integration works, step by step
Vantegrate handles the setup and rollout. Your data team authorizes, validates and starts using it.
Connection through the REST API
We connect to the Tableau REST API using a personal access token (PAT) at sign-in. On Tableau Cloud with MFA, the PAT is the required method. No scrapers and no fragile bridges.
Data sources and permissions
We define which published data sources the agent queries and with which Tableau role, so access to the data and the metadata respects the permissions you already have in place.
Querying through the VizQL Data Service
We set up the VizQL Data Service queries to pull the governed data without a viz, and the Metadata API for the catalog and the lineage the agent cites in every answer.
Validation with real data
We test with your data sources and your real questions, validate with your data and IT teams, and roll out in phases.
The VizQL Data Service vs. exporting a report
Tableau exposes several APIs on the same authentication foundation. The REST API manages site content (data sources, workbooks, views, permissions), the VizQL Data Service queries the data in a published data source without needing a visualization (with a JSON request that filters, aggregates and transforms), and the Metadata API resolves metadata and lineage. That combination is what lets you work with your governed data from outside Tableau.
The difference from exporting a report shows up in every question. The agent queries the current data in your data source at that moment, reusing the metric your team already validated, instead of working on a downloaded spreadsheet that's out of date. Because it points to the governed source, the number it returns is the same one the official dashboard shows.
| Dimension | Exporting a report | Vantegrate integration via the VizQL Data Service |
|---|---|---|
| Data source | A downloaded spreadsheet that drifts | Published data source, governed data |
| How you ask | Open the file and filter by hand | In natural language; the agent turns it into a query |
| Needs a viz | Yes, someone builds the report | No, the VizQL Data Service pulls the data without a viz |
| Metrics | Recalculated elsewhere, may differ | The same ones that feed your dashboards |
| Permissions and lineage | Lost when you export | Respects the role; lineage comes from the Metadata API |
Comparison of patterns; the exact scope is defined with your team based on your site, your data sources and your permissions.
Your data stays on your site: the agent queries your Tableau through the API; it doesn't take your database anywhere else. The connection uses a revocable personal access token and respects each user's Tableau role, so nobody sees data over chat that they couldn't see in a dashboard.
How the agent respects Tableau's permissions and limits
The integration respects Tableau's permissions. The Metadata API uses the same View capability as Tableau Server, and the user's permissions determine which metadata and content are visible; when the user lacks View permission on an asset, the Metadata API obfuscates that metadata by default, which is expected and is accounted for in the design.
Authentication also sets limits to keep in mind for automations. A personal access token expires if it isn't used for fifteen consecutive days (and on Tableau Server, by default, after a year), its secret is shown only once when it's created, and it doesn't allow concurrent sessions. That's why the integration is designed with dedicated tokens and retries, within your site's request limits.
About the API limits: Tableau applies request limits per site, and the Metadata API can return partial results on large queries. The integration is designed within those limits, with scoped queries and retries, and it works the same on Tableau Cloud and Tableau Server (the host and the site change, not the flow).
You've seen how it connects. Want to walk through your case?
Tell us how your Tableau is set up and we'll tell you what data we need and where the agent connects.
An agent connected to your Tableau vs. a standalone tool
| Standalone AI tool | Agent connected to your Tableau | |
|---|---|---|
| Data source | A copy that drifts | Your source of truth, live |
| Where your data lives | In a third-party system | In your environment, encrypted end to end |
| Result of each question | A one-off export | An answer from live data, where your team works |
| Permissions and audit | Broad and hard to audit | Read-only role and a log of every query |
| Reaching production | Often stalls as a pilot | Validated, phased rollout |
Why add an agent that answers from your BI
Third-party figures, each with its published source, to size up the context. None of them is a Vantegrate result.
Salesforce
Tableau is Salesforce's analytics and business intelligence platform
95%
Of generative AI pilots never reach production with measurable business impact
The Salesforce fact describes which ecosystem Tableau belongs to, not a Vantegrate result. The MIT NANDA figure reflects the gap between experimenting with AI agents and running them in production: adding the agent on top of your governed data, instead of rebuilding it, is what narrows that gap.
Which agent runs on your Tableau
The Suite's data product uses this integration through the official APIs: it queries your site's governed data when the question is asked and explains it where your team needs it.
The integration doesn't take your data anywhere
The principle is simple: AI comes to your data, not your data to the AI. The agent queries your Tableau through the official APIs and respects each user's role, and the agents run on certified cloud infrastructure.
- Connection through the official Tableau APIs (REST and the VizQL Data Service) with a revocable personal access token, with no scrapers and no fragile bridges.
- The agent queries your published data sources and their lineage; your governed data stays in your Tableau Cloud or Tableau Server.
- Access scoped by each user's Tableau role and minimum permissions, so nobody sees data over chat that they wouldn't see in a dashboard.
- The agents run on Salesforce or Oracle Cloud Infrastructure, whose SOC 2 and ISO 27001 certifications belong to those platforms, not to Vantegrate or Tableau.
Frequently asked questions about the Tableau integration
What data and IT teams usually ask before connecting an AI agent to their BI platform.
How do I integrate Tableau with an AI agent?
How do I integrate Tableau with an AI agent?
With a connection through the official Tableau APIs: the REST API for site content, the VizQL Data Service to query the data in a published data source without a viz, and the Metadata API for lineage. The agent turns your natural-language questions into those queries and gives you the figure with its context. Vantegrate handles the setup and rollout; your data team authorizes, validates and starts using it. Meet the data agent at Metrix.
Can you query the data without a dashboard already built?
Can you query the data without a dashboard already built?
Yes. The VizQL Data Service pulls the data from a published data source with a JSON request that filters, aggregates and transforms, without needing a visualization. That's why the agent can answer a specific question or combine several data sources without anyone having to create a new workbook or view in Tableau.
How does the integration authenticate, and does it work with Cloud and Server?
How does the integration authenticate, and does it work with Cloud and Server?
With a personal access token (PAT) at REST API sign-in; on Tableau Cloud with MFA enabled, the PAT is the required method. The authentication and API flow is the same on Tableau Cloud and Tableau Server; only the host and the site change. The PAT expires after a period of inactivity and, on Tableau Server, after a year by default, so automations are designed with dedicated tokens.
Does the integration respect Tableau's permissions?
Does the integration respect Tableau's permissions?
Yes. The Metadata API uses the same View capability as Tableau Server, so the user's permissions determine which metadata and content are visible, and metadata the user can't view comes back obfuscated. The agent never shows data over chat that the user couldn't see in a dashboard.
Metrix, the data agent
How the team asks in plain English and gets answers from the real data, without static reports.
Learn moreAll integrations
The full map: the BI tools, ERPs, CRMs and other tools the AI agents connect to.
Learn moreAI agents for enterprises
The map of the Vantegrate suite: sales, marketing, documents, data and logistics.
Learn moreLet's connect AI agents to your Tableau
Tell us which data sources you have published and we'll show you how the agent queries them, which permissions we request and what asking your BI questions in natural language looks like. A 30-minute conversation, no commitment.
Francisco Morales, co-founder, takes your call. We reply on WhatsApp within 4 business hours, no strings attached.
