AI in BI: Why Most Tools Miss the Mark (And What We Can Do About It)
Over the last six months, I’ve had a growing number of conversations with clients and prospects around one big question:
“How can we actually use AI to make our reporting smarter?”
Some were just curious about plugging AI into their dashboards to look more cutting-edge. Others had genuine hopes of extracting better insights from their data using AI.
And truthfully? At first, I didn’t have a clear solution I could confidently recommend.
That didn’t sit well with me.
As a Business Intelligence consultant, my job is to help people make better decisions with data. So I decided to dig deeper and find a way for AI to actually move the needle in real-world BI environments.
The Real Problem with AI in BI Tools
AI is showing up everywhere — especially in BI platforms. But in my experience, most of what’s labeled “AI” in dashboards today is just automation with lipstick.
It tells you:
“Revenue is up 10%.”
“Product X had the highest sales.”
“Traffic peaked on Tuesday.”
But that’s not insight — that’s just observation.
It’s not why something happened.
It’s not what to do next.
It’s certainly not unexpected.
And that’s the problem. Current AI features in most BI tools are limited by:
- Predefined prompts and visuals
- Surface-level summaries
- No context, no curiosity, and no direction
It’s all reactive. And business leaders are hungry for more.
What AI Could Be Doing Instead
We know AI can do more than read back numbers.
It can reason, explore, even hypothesize — when it’s allowed to.
So I asked:
- What if we removed the dashboard as a limitation?
- What if we gave AI direct access to raw data?
- What if it could deliver tailored summaries, adjusted to any business question?
And that’s how the BI Summarizer Framework (BISF) was born.
The BI Summarizer Framework (BISF)
After months of iteration, I developed a lightweight, flexible framework designed to unlock real AI-driven analysis, not just surface-level summaries.
Here’s how it works:
- Data is pulled from a warehouse or analytics platform via SQL.
- A custom script wraps that query with context and a tailored prompt.
- The request is passed to Gemini, chosen for its ability to work with large structured datasets and provide deeper insights.
- Gemini returns a contextual summary — complete with trends, risks, patterns, and even recommendations.
- That summary is sent to Google Sheets, Google Docs, email, or Slack. Wherever decision-makers are.
This process is BI tool agnostic, runs automatically, and adapts to any dataset or use case.
You can run it daily, weekly, or monthly, giving teams fresh, digestible insights without anyone lifting a finger.
It’s Not Replacing Dashboards — It’s Augmenting Them
Dashboards still matter. But they’re passive. You have to interpret them, know what to look for, dig deeper, explain it to someone else.
BISF flips that on its head.
It surfaces the insight directly. No clicks. No filters. No guesswork. Just tailored summaries, ready to act on.
So Where Does This Go From Here?
I now use this framework with clients — and teach it inside my BI Consultant Masterclass and BI Analyst Starter Program.
It’s become the answer I was looking for when those early AI questions started rolling in.
And now, when someone asks how to incorporate AI into their BI workflow, I don’t have to hesitate.
I’ve got a solution that actually works — and I’m excited to share it.
👉 Learn more at: bisummarizer.com
📘 Or check out my courses at: learnbi.academy
