Introducing text-to-dashboard

BI as we know it is dying.

That's not a provocation. It's just where things are heading. In a few years, we'll look back and find it hard to believe that building a dashboard required hours of manual work: dragging charts, wiring up filters, formatting KPIs, and then doing it all again every time something changed.

Today we're launching generative dashboards in Fabi. Anyone can now build a fully interactive dashboard by describing what they want, on top of secure, trusted, centralized data. No chart builder. No extra steps.

This post is also about where BI is going, because we think the shift happening right now is bigger than any single feature.

The viz layer is getting completely reimagined

For decades, the visualization layer in BI has looked roughly the same. You connect a data source, you drag fields into a chart builder, you configure axes and filters, and eventually you publish something. The process is manual, the output is static, and the moment your underlying data or business changes, you're back to square one.

AI is changing all of that. The chart builder as a concept is becoming obsolete. The question is no longer "how do I configure this visualization?" It's "what do I want to see?" and the tool figures out the rest.

This is what we mean by generative BI. You describe the dashboard you need, and it gets built. With Fabi, you get:

• KPIs, filters, dropdowns, and executive summaries generated from a single prompt
• Full customization with follow-up prompts — change your brand colors, swap a chart type, add a new filter
• Near-zero build cost, so you can prototype freely and spin up new views on demand
• One-click publish, hosted and managed by Fabi

Dashboards are becoming fluid. They shouldn't be artifacts you plan in advance and maintain by hand. They should exist when you need them and reflect your business as it actually is right now.

Good data foundations make generative BI work

Generative dashboards are only as good as what sits underneath them.

Dashboards run on large, messy, business-critical data. If your metric definitions are inconsistent, if different teams are pulling revenue numbers from different places, the output will reflect that. The AI needs something solid to work from.

That means:

• Centralized, governed data connections across your warehouse and applications
• A shared data model that the whole team works from
• Metric definitions and business logic that live in one place and stay consistent across every dashboard, every analysis, every report

When you connect your data warehouse to Fabi, you're not just connecting tables. You're bringing in context: your definitions, your relationships, your business logic. The Analyst Agent works from that context every time, which means every dashboard it generates reflects how your business actually works, not just how your schema happens to be structured.

Context by role matters too

Different people in your organization need to see different things. A CFO and a growth marketer might both want a dashboard, but the metrics they care about, the level of detail they need, and the data they should have access to are completely different.

Generative BI needs to understand that. Context management goes beyond metric definitions. It covers role-based access, relevant questions per team, and surfacing the right information to the right person without them having to know exactly where to look. This is core to what makes self-service analytics actually work at scale.

This is part of the longer arc of what we're building. BI on demand, for everyone in your organization, grounded in data they can actually trust.

What's new in today's launch

For those who've used Fabi before, here's what changed. Analysis and dashboard used to be two separate steps. You'd run your analysis in a Smartbook, and then if you wanted a shareable dashboard you had to go build it separately. That step is gone. It's now one fluid motion from question to published dashboard.

The AI can now generate full dashboards directly, complete with filters, inputs, dropdowns, KPI cards, and executive summaries. Everything is customizable with a follow-up prompt. We handle hosting, caching, pagination, and context management for large and messy datasets so the output is reliable even on real-world data that isn't clean or simple.

Publish in one click. Share with your team. Update it with a prompt when something changes.

Welcome to the future of BI

We're still early in this shift. Generative dashboards are a meaningful step, but the longer vision is a world where any person in your organization can get a reliable, accurate, on-demand view of the data that matters to them, without waiting on a data team or learning a new tool.

That requires good data foundations, thoughtful context management, and AI that's built specifically for the complexity of real business data.

We're building toward that. Today is one step.

Try Fabi free and build your first generative dashboard in minutes.