Fabi 2.0 - The AI data analyst for all your data

When Lei and I founded Fabi, we had one simple vision: everyone from product managers to growth marketers should be able to pull together their own analysis. Data analysis should be a team sport.

This came from personal experience. When we were working together, I was constantly copying and pasting SQL queries or exporting data to analyze product usage, revenue numbers, or troubleshoot customer issues. Lei, as a data scientist, was constantly having to help me. I wasn't completely illiterate in SQL or Python—I understood how they worked and had coded plenty in the past—but I simply couldn't keep up with the pace the business demanded.

When GPT-3 came out, the solution seemed obvious. We set out to create more than just a text-to-SQL chatbot. We wanted to build an AI data analyst that truly understood your business, in a collaborative environment where teams could work together seamlessly.

Since launching in early 2025, we've worked with hundreds of product managers, founders, growth marketers, and data teams. They use Fabi for ad hoc analysis, shipping dashboards faster, and collaborating on insights across teams. But there was a challenge: not all data was centralized. Many teams weren't ready to invest in ingesting and modeling all their data sources, but still needed answers. Your Google Ads data might not be in Snowflake, but you still want to optimize your keyword strategy. Or you're a small team using TikTok without access to a data team. Those insights remained out of reach.

The shift to AI-native business intelligence

We're at an inflection point in how companies work with data. The traditional BI stack—built around drag-and-drop dashboards and pre-built visualizations—was designed for a world where data teams were bottlenecks and business users needed templates.

That world is ending.

AI business intelligence tools are fundamentally changing what's possible and Fabi is at the vanguard. Instead of waiting for a dashboard to be built, teams can ask questions in natural language and get answers immediately. Instead of being constrained by pre-defined metrics, self-service analytics means anyone can explore their data freely. Instead of static charts, AI data visualization adapts to the question you're actually asking.

We believe that in 12-24 months, the current approach to BI and data visualization will feel archaic. The future isn't about building more dashboards—it's about having a conversation with your data. And critically, it's about data collaboration: everyone on the team working together in the same environment, building on each other's analysis in real-time.

That's the world we're building at Fabi.

What's new in Fabi 2.0: Your AI data analyst for all your data

We're introducing connectors to popular data sources. Now you can connect not just databases (Postgres), data warehouses (Snowflake, BigQuery, Databricks), and files (CSV, Excel, Google Sheets)—but also applications like HubSpot, Stripe, PostHog, Amplitude, Shopify, and Google Ads.

When you connect your data, Fabi analyzes the metadata and structure to help the AI provide accurate answers. For certain data sources, Fabi even automatically builds cleaned-up data models under the hood, so the AI has a real understanding of your data within minutes of connecting.

This helps teams that have data locked up in different applications and spend days wrangling spreadsheets.

We're also introducing an AI context configuration layer. In addition to automatically learning your business, you can give the AI additional information to help it provide better answers.

Where we go from here: The future of AI for business intelligence

We're just getting started. Today, we offer a comprehensive AI data analysis platform with a wide set of connectors, but there's still a lot of work to do.

We're constantly improving the AI—both the agentic framework and the data layer—to ensure it provides fast, accurate answers without requiring expertise in data modeling or BI.

We're also investing in tools that aren't necessarily AI-related to make data analysis and collaboration easier. We know AI isn't the right answer for every problem. Sometimes you just want to do it yourself. So we're continuing to build power-user tools for technical teams to make the possibilities in Fabi truly expansive.

Thank you for being on this journey with us. Everything we build is for you.

Marc, Lei, and the rest of the Fabi team