Tableau MCP Integration

Weldable's Tableau MCP integration gives your AI agents access to your Tableau dashboards, data sources, and workbooks through natural language. Your agent can query published data, pull metrics from Tableau Pulse, retrieve dashboard information, and manage workbooks without anyone logging into Tableau Server or Tableau Cloud.

Tableau has invested heavily in its own MCP support, releasing open-source developer primitives that cover the REST API, Metadata API, VizQL Data Service, and Pulse API. Weldable connects to these capabilities through a single authenticated interface, so your agent can pull data from Tableau alongside Slack, Google Sheets, and any other connected service.

Use cases

On-demand metric lookups

Your agent answers questions like "what was last month's revenue by region?" by querying the underlying Tableau published data source through the VizQL Data Service API. It pulls the data, formats it into a readable summary, and delivers the answer in Slack or whatever channel you're working in. Business users get answers in seconds without opening a dashboard or remembering which workbook has the data they need.

Automated KPI reporting

Your agent pulls key metrics from Tableau Pulse (subscriptions, metric definitions, and value bundles) and compiles them into a weekly report. It compares current values against the previous period, flags metrics that moved more than 10%, and posts the summary to your leadership channel. The report includes specific callouts for metrics that crossed alert thresholds.

Dashboard inventory management

Your agent queries the Tableau REST API to list all workbooks, dashboards, and data sources on your server. It identifies dashboards that haven't been viewed in 90 days, data sources with broken connections, and workbooks owned by deactivated users. The audit report helps your Tableau admin clean up stale content and reclaim server resources.

Data-driven Slack alerts

When a metric in Tableau crosses a defined threshold, your agent detects the change during its regular polling and sends a targeted alert to the relevant team. If customer churn rate exceeds 5%, the agent pulls the supporting data from Tableau and posts it to the customer success channel with the trend line and top contributing segments. The alert includes enough context for the team to act immediately.

Cross-platform data assembly

Your agent pulls financial data from Tableau, combines it with project status from Google Sheets and team updates from Slack, and produces a consolidated executive briefing. Each data source contributes its piece, and the agent handles the assembly. The output is a single document or message that would normally take an analyst an hour to compile manually.

How it works

Connect your Tableau Server or Tableau Cloud account through Weldable. Authentication uses personal access tokens or site-level credentials, and Weldable manages the session lifecycle. Your agent authenticates once and maintains access across requests.

Once connected, describe what you need in plain English. Say "show me the top 10 products by revenue from the sales dashboard" or "list all workbooks last updated more than 6 months ago" and Weldable translates your intent into the appropriate Tableau API calls. Results come back structured and ready for your agent to summarize, compare, or forward to another tool.

Tips

Tableau has multiple APIs for different purposes. The REST API manages content (workbooks, projects, users). The Metadata API returns column descriptions and lineage. The VizQL Data Service runs queries against published data sources. Your agent selects the right API based on your request, but understanding the separation helps when debugging unexpected results.

Tableau Pulse metrics are separate from dashboard data. Pulse is Tableau's AI-powered metrics layer. If you've set up Pulse definitions, your agent can query them directly for metric values and insights. If you haven't, the agent falls back to querying the underlying data source.

Published data sources are the best query target. Querying a published, certified data source gives your agent access to clean, governed data with proper column descriptions. Querying embedded data sources in individual workbooks often returns less structured results.

Content permissions in Tableau are project-based. Your agent can only access workbooks and data sources that the authenticated user has permission to view. If your agent returns empty results for a query you expected to work, check the project permissions on the Tableau side.