Introducing ThoughtSpot’s Agentic MCP Server

Make your AI agents data-savvy experts

The AI agent revolution is transforming how we work, but most analytics platforms are stuck in the past—forcing you to context-switch between your agents and separate BI tools to get data insights. 

This fragmented approach creates friction, breaks workflows, and ultimately slows down decision-making. When speed matters most, you need your AI agents to seamlessly access and analyze enterprise data without the traditional barriers of maintaining custom integrations and limited API functionality.

That's why we're excited to announce the launch of ThoughtSpot's Agentic Model Context Protocol (MCP) Server, now generally available for all customers. ThoughtSpot is the first major BI analytics platform to provide an MCP server with native natural language to data capabilities to any agent or business user in the market.  Now, you can effortlessly bring the full power of ThoughtSpot's agentic analytics directly into any AI agent, application, or platform that supports MCP—eliminating context switching and delivering insights exactly where your teams work, including Claude, ChatGPT, Gemini, or any AI agent, application, or platform that supports MCP.

With ThoughtSpot’s Agentic MCP Server, anyone can ask complex questions about their data using natural language within their preferred AI environments and receive comprehensive, actionable insights—complete with summaries, AI-augmented dashboards, and intelligent follow-up suggestions. 

For developers and product teams, this means dramatically reduced integration complexity and costs, with the ability to build once and deploy everywhere across the growing MCP ecosystem.

AI that thinks ahead and follows through

Unlike traditional BI integrations that simply translate questions into SQL and return static results, ThoughtSpot’s Agentic MCP Server harnesses the full analytical reasoning power of Spotter, your AI analyst

When you ask a question through your agent, you're not just getting a basic API response—you're accessing Spotter's sophisticated analytical skills that can reason through complex business problems, automatically generate relevant follow-up questions, and provide comprehensive insights that drive action.

With ThoughtSpot’s Agentic MCP Server, your agent doesn't just answer your immediate question; it anticipates what your team might need next. It leverages Spotter skills to automatically identify related metrics worth exploring and ask follow-up questions to unlock valuable context. 

Your agent can even combine multiple insights into interactive Liveboards with auto-generated summary notes, empowering you to save insights, share with teams, and further explore relevant data by asking Spotter about any Liveboard visual. You can change the look and feel of your Liveboard notes simply by describing these changes to your agent.

Now you can infuse critical capabilities, transforming your AI agent into a true analytical partner capable of conducting thorough, multi-faceted analysis—not just a query translator.  With ThoughtSpot’s Agentic MCP Server, your AI agent mirrors how expert analysts think, building on initial findings, exploring correlations, and presenting insights in the most actionable format possible.

The most comprehensive suite of agentic skills

While other BI vendors offer basic API access with limited functionality, ThoughtSpot’s Agentic MCP Server delivers the market's most complete set of analytical capabilities designed specifically for AI agents. Our comprehensive suite includes intelligent question generation, advanced answer retrieval, and AI-augmented dashboard creation—all working seamlessly to provide a complete analytics experience.

Here’s how it works: 

  • Your agent gains access to our sophisticated question generation engine that can take high-level queries and automatically break them down into specific, answerable analytical questions based on your available data sources. 

  • The answer retrieval system then executes these questions against your ThoughtSpot data sources, returning rich insights with full context. 

  • Finally, the Liveboard creation capability allows your agent to automatically combine related insights into interactive Liveboards that stakeholders can continue exploring. It also allows you to provide design directions to your agent, so you can customize the look and feel of your Liveboard summaries.

This isn't just a collection of disconnected API endpoints—it's a thoughtfully designed analytical toolkit where each capability enhances and builds upon the others. The question generation informs better answers, the answers provide material for richer dashboards, and the entire system learns from your data structure, resulting in increasingly relevant insights. 

With new analytics skills continuously being added to the platform, your agent's capabilities will only grow more powerful over time.

Purpose-built for natural language queries

Most BI vendors treat natural language as an afterthought, forcing your agent to translate your questions into generic API calls. This means the agent can’t have knowledge about the schema of your data, the terms you use in your business, and the data itself, which could contain billions of rows. 

This creates friction in the interpretation layer where nuanced business questions get lost in translation, resulting in misunderstood queries, irrelevant results, and frustrated users. ThoughtSpot’s Agentic MCP Server takes a fundamentally different approach with native natural language support.

Natural language querying is built into ThoughtSpot's core architecture, not bolted on. Our technology has been proven and refined through millions of real-world business queries, ensuring trust and accuracy through our: 

When your agent sends a question to ThoughtSpot’s Agentic MCP Server, it leverages our advanced natural language capabilities, providing the intelligence required to handle complex, context-rich questions. 

Whether you're asking, "How can we improve on quarterly revenue trends excluding one-time charges?" or "What were customer churn patterns in the Northeast region compared to Q3 last year?" ThoughtSpot’s Agentic MCP Server understands the business context and nuance behind your questions. 

The result: Insights that actually address what you're trying to understand—not just what a generic agent thinks you might want.

Ground your AI agents in your business reality

The biggest challenge facing AI implementations in analytics isn't technical—it's ensuring that AI-generated insights actually reflect your business reality. Generic AI models interpret general patterns, but they don't understand your specific metrics, business logic, or organizational context. ThoughtSpot’s Agentic MCP Server solves this obstacle by grounding all AI insights in our industry-leading agentic semantic layer. We call this Data Augmented Generation (DAG).

Every insight generated through ThoughtSpot’s Agentic MCP Server is anchored to your organization's specific business definitions, logic, terminology, and, of course, facts. Your AI agent doesn't just access raw data—it analyzes data enriched with your institutional knowledge about what metrics mean, how they should be calculated, and what context makes them meaningful for your business decisions.

This semantic grounding means you can trust the insights your agent provides, knowing they are built on the same reliable foundation that powers your existing analytics workflows. Whether it's complex revenue recognition rules, custom KPI definitions, or industry-specific calculations, your AI agents work with data that reflects how your business actually operates, not generic assumptions about how it might work.

Enterprise-grade agentic skills are available today

While competitors are still experimenting with basic MCP implementations or haven't officially launched their servers, ThoughtSpot’s Agentic MCP Server delivers production-grade analytical capabilities that are ready for production use today with the same reliability and security standards that enterprise customers depend on. You can use ThoughtSpot’s Agentic MCP Server today on MCP-compatible platforms like Claude and Gemini.

For those who are interested in integrations with ChatGPT, ThoughtSpot’s integration with OpenAI responses API is available in GA, and the integration with OpenAI Deep Research is in early preview.

MCP ThoughtSpot ChatGPT answer

Governance and security

ThoughtSpot’s Agentic MCP Server uses the OAuth 2.1 protocol, ensuring that all the permissions and access controls you've already established in ThoughtSpot are automatically respected when your agents access data. This means your governance policies, user permissions, access, and security protocols remain intact while enabling powerful new AI-driven workflows.

Unlike other vendors who are still working through the technical challenges of MCP integration, ThoughtSpot has solved the complex problems of making analytical AI both powerful and enterprise-secure. 

For example, instead of requiring developers to install Python or Node.js at a specific version—which may be inconvenient and pose security risks—ThoughtSpot’s Agentic MCP Server is securely hosted and accessible with just a URL. Your teams can start building with confidence, knowing they're working with proven technology that scales to meet enterprise demands.

Turn your AI agents into analytical powerhouses

The value is clear: trusted, business-contextual analytics delivered exactly where your teams need them, with the full reasoning power of agentic AI, and the security and governance your enterprise requires. Whether you're looking to enhance existing AI implementations or build new agentic workflows, ThoughtSpot’s Agentic MCP Server provides the analytical foundation you need.

ThoughtSpot’s Agentic MCP Server is available today at no additional cost for all ThoughtSpot customers with Spotter licenses. Access our public GitHub repository today for immediate implementation.

Not an existing ThoughtSpot customer? Schedule a demo today to see how ThoughtSpot’s Agentic MCP Server can revolutionize your team's analytical workflows, or start your free trial to experience the future of agentic analytics firsthand.