Fabi.ai Product Updates

Ask questions and get answers right in Slack

Author

Marc Dupuis
February 17, 2026

Get answers right in Slack

Your team's busiest channel just got smarter. Instead of waiting for data team availability or context-switching to another tool, you can now ask Fabi questions directly in Slack, and get analysis back in seconds.

Early adopters are seeing real impact: organizations using the Slack app are making 5X more data requests per week. Not because they suddenly need more data, but because asking is finally frictionless.

What you can do with our Slack app:

  • Ask questions in natural language ("What were our top performing campaigns last month?")

  • Get instant charts, tables, and insights without leaving Slack

  • Share results with your team in the same thread

  • Follow up with clarifying questions just like you would with a colleague

Getting started takes 3 minutes:

  1. Connect your data sources to Fabi (databases, warehouses, CRM, marketing tools, whatever you use)

  2. Give the AI a bit of context about your data (optional but helpful)

  3. Install the Fabi Slack app and start asking questions

Try it with a simple question first, you might be surprised how quickly you get a useful answer.

More AI capabilities: Cleanup and reorganize

The AI Analyst can now help tidy up your work. Need to merge duplicate analyses, split a complex notebook into parts, or delete outdated content? Just ask.

A lot more coming on this front very soon. Stay tuned.

Quality of life improvements

  • Auto-organization: New ad hoc analyses go into an "Ad hoc" folder automatically; Slack queries get their own folder. Less clutter, easier to find what you need.

  • KPI widgets: The AI can now generate KPI cards much more reliably 

Other reads and ideas

The end of legacy BI?

Tableau, PowerBI, and Looker are still deeply embedded in most organizations. But for how much longer?

We're at a point where AI can generate dashboards and reports in a fraction of the time these platforms require. It's not hard to imagine that within 24-36 months, most BI interactions happen through natural language rather than point-and-click interfaces. Legacy tools simply weren't designed for this.

Our co-founder Marc recently walked through a concrete example of what this shift looks like and why it matters.

Clawdbot and the agent economy

Even if you're not on X, you've probably heard rumblings about Clawdbot and Moltbook, a social network for AI bots.

Worth understanding even if practical applications are still emerging. A few takeaways:

  • AI agents are powerful but risky if not used in controlled, trusted environments. Running agents locally doesn't automatically make them safer, it can actually increase risk if you don't know what you're doing.

  • Agents are increasingly autonomous. It's clearly a matter of time before they work together independently. The advantage will go to teams that learn to harness this multiplier effect.

Jon Krohn breaks it down well in this LinkedIn post.