• Daily Success Snacks
  • Posts
  • The Funniest Part About “AI-Powered Analytics”? The Data Model Is Still Broken.

The Funniest Part About “AI-Powered Analytics”? The Data Model Is Still Broken.

Before adding AI copilots and predictive dashboards, somebody should probably fix the circular references first.

Author

Glenda Carnate
May 22, 2026

Read time: 2.5 minutes

No matter how hard I try to get our company to address the actual problem in every analytics meeting, it keeps getting ignored.

Leadership: “We need analytics that are powered by AI.” Response from room: “We need to integrate AI into Qlik!”

Meanwhile, the developer remaking Qlik is dealing with:

  • synthetic keys

  • circular references

  • duplicate fields

  • an overall data model held together by sheer will.

What does a Qlik Developer need?

1. Clear Data Model
Before using AI features:

  • Resolve synthetic keys

  • Remove unnecessary joins

  • Eliminate circular references

  • Dismiss unwanted tables

Bad relationships = Bad insights

2. Standardized Business Reasoning
When each dashboard calculates KPIs its own way, you can’t depend on any AI-powered summary… it’s unreliable from the beginning.

Therefore, standardization is more significant than hype.

3. Fewer separate Data Sources
Most reporting chaos begins with teams pulling data from:

  • Spreadsheets

  • Legacy Systems

  • Manual Exports

  • Shadow Databases

A single reliable data pipeline is worth 10 or more “Smart” Integrations.

4. Analytics Users Can Trust
Most BI Issues are not with visualization… they are with credibility.

When users stop trusting the numbers, adoption stops, no matter how advanced the platform is.

💡Key Takeaway: 

AI can restate dashboards, but cannot heal corrupted data relationships. Before adding the "AI-powered Analytics" label, ensure your foundation is not powered by confusion.

👉 LIKE if you have reallocated your time from building dashboards to fixing synthetic key issues.

👉 SUBSCRIBE now for more insightful commentary on Qlik, BI, analytics, AI and Data Strategies.

👉 Follow Glenda Carnate for a weekly humorous breakdown of every data team's problems that they will not tell you about.

👉 COMMENT with the worst circular reference nightmare you’ve dealt with.

👉 SHARE this with others who think that AI can resolve a broken data model.

Reply

or to participate.