Build faster, act smarter: Introducing agentic AI with the assistant, MCP server capabilities, and flexible LLM options

Actionable intelligence for everyone

With this release, Sisense is introducing several updates to our agentic AI for analytics. At the heart of this milestone are three major announcements:

• The assistant, now generally available, brings natural language analytics to every builder and user. 
• MCP (Model Context Protocol), a new connection layer, allows external AI agents such as ChatGPT or Claude to help you build visual analytics (charts, metrics, dashboards) in their own interfaces.  Our product is now ready for you to work with MCP.
• Flexible LLM options: Customers can connect their own LLM, and a Sisense-managed service is also available.  The offering provides all users with flexible AI solutions.

These advancements make it more practical for teams to build and embed analytics into their products every day. They shorten the path from question to action, helping product leaders, developers, and data teams use Analytics platform as a service (AnPaaS) to build, scale, and differentiate faster.

Plus, end-users will love the ability to self-serve insights intuitively, right in the app where they’re already working.

Assistant: From prompt to production

Assistant, one powerful feature of Sisense intelligence, is now fully available for all Sisense customers. Assistant is the primary way teams experience Sisense Intelligence. It gives you a faster, more intuitive way to create and manage analytics.

With assistant, you can use  natural language to:

• Spin up a data model with generated sample data for quick prototyping
• Create charts and iterate on them in conversation
• Assemble and refine full dashboards

From there, continue to explore, edit, and manage your charts and dashboards in conversation, while all data remains secure and consistent with your organization’s models.

Why it matters

With assistant, no SQL or data modeling expertise is required for anyone building and embedding analytics. 

Specifically:

• For product managers, the assistant speeds up prototyping and eliminates long back‑and‑forth cycles with technical teams.
• For developers, it reduces repetitive dashboard requests so they can focus on advanced integrations.
• For end-users, it enables self‑service exploration.

The result is a smoother analytics workflow that accelerates insight delivery, no matter a person’s technical background.

MCP: Connect your governed data to external AI agents

Model Context Protocol (MCP) is a secure way to connect your Sisense data to external AI tools like ChatGPT, Claude, or Snowflake Cortex. And the Sisense product is now ready for you to work with MCP, should you choose to do so. MCP is how you meet your users where they are while keeping your governance intact.

With MCP, you can:

• Allow users to ask questions directly inside AI tools they already use.
• Ensure those answers come from your governed semantic models.

MCP allows you to pull external AI capabilities into Sisense itself, creating a truly open and extensible analytics framework. Learn more.

Why it matters

• For developers, MCP extends analytics experiences into new interfaces without rebuilding from scratch.
• For data leaders, it preserves data lineage, governance, and context, even when external tools are involved.
• For product teams, it allows analytics to appear exactly where users are already working, inside familiar environments. This drives faster adoption and higher engagement.

Further MCP functionality is on our roadmap.  To help shape Sisense product updates, we’d love to hear from you on how important MCP is to your team, your product roadmap, and the agents you’d want Sisense to interact with.

Explore how real-time AI insights are transforming embedded analytics. Learn about Sisense Intelligence.

Read more

Flexible LLM options

Sisense currently supports bringing your own LLM (OpenAI) and will support additional tools, including Amazon Bedrock and GCP later this year.

In addition, the Sisense-managed LLM service is now available to select customers in private preview and will be available to all customers in Spring 2026. At that time, you will be able to chose between bringing your own LLM or using the Sisense managed service.  The Sisense-manged offering will provide a streamlined way to power assistant, semantic enrichment, and narrative features without requiring teams to manage their own LLM infrastructure.

As we get closer to general availability (GA) launch, we’ll share how teams can evaluate the service.

Agentic AI: Intelligent decision making

Sisense is building toward a composable analytics ecosystem powered by intelligent agents. These modular AI features are designed to help you explore, generate, and interact with data in smarter ways. 

These agents serve as the foundation for assistant, and are accessible to you:

1. Inside the Sisense platform: Agents power the assistant, enabling you to use natural language to create dashboards and get insights. On the roadmap: agents available in our SDK. 
2. Outside the Sisense platform: If you prefer, build your own assistant and embed our agents via Sisense APIs or MCP.

This approach supports a new standard for embedded analytics; one that’s flexible, governed, and infused with intelligence across your workflow.

Other highlights in this release

Along with the assistant GA and MCP debut, this release includes additional improvements designed to enhance both performance and usability:

• Semantic enrichment upgrades, including human‑in‑the‑loop editing.
• Faster query performance and access to multiple chat sessions for multi‑step workflows.

These updates make it easier to maintain a consistent, high‑quality analytics environment while taking full advantage of new AI capabilities.

What’s next

This launch sets the foundation for what’s coming in 2026. Upcoming milestones include:

• Assistant support for data modelling, simplifying schema creation and maintenance.
• Expanded MCP integrations with additional AI partners and enterprise data services.

Sisense is building toward an analytics platform that doesn’t just answer questions but helps you design, build, and act on insight, all within one governed environment. To learn more about our updated features and offerings, visit Sisense Intelligence: AI features for modern analytics.

Do you have more questions? Contact us.

Start your free trial

Get a full-featured 7-day trial with guided sample data, or connect your own.

Try free