Sisense product roundup: Smarter insights and AI-first workflows with release 2025.4

This quarter, we focused on helping builders of all types to create analytics experiences that are faster, smarter, and easier to maintain. From powerful enhancements to the assistant to features that simplify dashboard design and model management, these updates are all about giving you more control and better outcomes with less effort.

In this post, we’ll focus on features being released in Sisense version 2025.4, plus relevant 2025.3 service pack updates, if there are any relevant changes in those releases. 

We’ll walk through key improvements to Sisense Intelligence, new features that streamline workflows and improve data context, and end with a sneak peek at what’s coming next.

Smarter assistant, better answers, more control

We’ve unified the assistant experience across Sisense, making it smarter and more contextual than ever. The assistant for creators (currently in Beta) lives in its own tab and supports every stage of the analytics lifecycle—from data modeling to dashboard design to embedding. The in-dashboard assistant (generally available), embedded directly into the Analytics tab and dashboards,  is now significantly upgraded with a streamlined interface and faster performance.

These improvements we’ve rolled out help developers and product teams build faster and get more relevant results, and empower users to explore data more intuitively using natural language.

Improved NLQ accuracy: We’ve refined the assistant’s natural language query (NLQ) engine to return better, more precise answers, helping you go from question to insight with fewer iterations.

Automatic semantic enrichment updates:  For every data model that you select to enrich, this feature allows you to add semantics. We’ve also introduced a human-in-the-loop workflow, so you can review, edit, or choose to remove any AI-generated descriptions before they go live.

More powerful widget creation flows: The assistant now enables you to manage dashboard filters, use new shortcuts, and have multiple chat sessions, as well as improved control over chat history. These updates are designed to fit your workflow without any disruption, giving you greater flexibility overall and making it easier to revisit previous queries or clear your session as needed.

The enhancements above reflect our commitment to delivering an assistant that’s not just helpful, but truly developer-friendly, with speed, transparency, and customizability built in.

Build smarter dashboards and keep your models AI-ready

The following updates help you deliver more relevant experiences to your end users, while simplifying the way you manage and maintain your data models behind the scenes.

Keep your semantic model fresh, without a rebuild

We’ve introduced the ability to publish semantic changes without triggering a full build. This makes it faster and easier to keep your models current—whether you’re updating display names, refining descriptions, or tweaking tags.

Now, instead of waiting on a rebuild, you can apply those changes instantly and focus on exploration and insights.

But this isn’t just a speed improvement; it’s also a quality boost. A well-maintained semantic layer helps your AI features perform better, especially when using the assistant or other generative tools. Clean, consistent metadata leads to clearer answers and smarter recommendations.

Looking ahead

We’re continuing to raise the bar for what’s possible with embedded, AI-powered analytics, and we’re already making progress on several major Q4 initiatives.

One of the most requested features is on the horizon: left join support. In the first phase, you’ll get model-level join control, giving you more flexibility and precision when defining relationships between tables. This will unlock new possibilities in data modeling and help teams shape their analytics to match real-world business logic.

We’re also making it easier to use AI across the Sisense platform by building out a Sisense-managed LLM service. This will allow you to power features like the assistant without needing to bring your own LLM key, thus simplifying setup while maintaining enterprise-grade control and security.

And because trust is foundational to every analytics experience, we’re adding email-based two-factor authentication (2FA) to strengthen account protection. This is part of a broader set of ongoing improvements aimed at making Sisense faster, more secure, and easier to use at scale.

We want to hear from you

Your feedback is crucial in shaping future releases, so be sure to voice your ideas and suggestions in the Product Feedback Forum. Stay tuned for more updates as we continue to push the boundaries of innovation together.

Curious about what you’ve seen? Check out the free trial

If you’re still exploring embedded analytics solutions and want to try Sisense for yourself, you may. Our new free trial experience lets you get hands-on with the platform, no strings attached. Or, to learn more about analytics solutions that will keep you competitive, schedule a demo.