Smarter, Fairer, and More Transparent AI in Qlik Predict™
In a previous blog post, we introduced one of the most significant advancements in Qlik Predict™ to date — multivariate time series (MVTS) forecasting, bringing enterprise-scale accuracy and context to complex prediction problems.
But MVTS is only part of the story. Over the past few months, we’ve continued to enhance Qlik Predict with several powerful updates designed to make predictive AI more trustworthy, fair, and connected for our customers.
This post highlights a few of those enhancements and how they expand Qlik Predict’s value across business and analytics workflows.
Model Training Report: Clearer Transparency for Every Model
Every prediction starts with trust. The new model training report creates a detailed PDF for each experiment so you can see exactly how a model was built.
The report covers:
Data used and preprocessing steps applied
Feature engineering decisions
Algorithms tested and hyperparameters selected
Performance metrics and evaluation results
Why it matters:
Provides full auditability for compliance requirements, including alignment with emerging regulations such as the EU AI Act
Builds stakeholder confidence with explainable machine learning
Supports both business users and data science teams by documenting every decision along the way
[Example model training report]
Automatic Class Imbalance Handling: Fair Predictions Without Extra Effort
Many real-world datasets are uneven, such as churn where 95% of customers stay and 5% leave. Without correction, models can ignore the minority class.
Qlik Predict now detects imbalance and applies corrective techniques automatically:
Improves accuracy on minority outcomes such as fraud cases or machine failures
Reduces the manual rebalancing work usually required
Ensures analysts and decision makers get reliable predictions without added complexity
Enhanced AI Connector: More Ways to Bring Data In
AI is only as good as the data behind it. The updated AI Connector makes it easier to integrate Qlik Predict with the services and data sources you rely on.
With broader connectivity and streamlined setup, teams can:
Combine external data with Qlik Predict for richer models
Bring in live signals to improve real-time forecasting
Expand predictive workflows into more parts of the business
Why These Updates Matter
Together, these improvements bring more transparency, fairness, and flexibility to predictive AI in Qlik:
Transparency with audit-ready model reports
Reliability with automatic handling of class imbalance
Connectivity with enhanced AI data integration
All three are available in Qlik Predict today.
Get Started
Log in to your Qlik Cloud tenant to try these new capabilities, or reach out to your Qlik representative to learn how predictive AI can drive measurable impact in your business.
With Qlik Predict, you get more than accurate forecasts. You get governed, explainable AI that works where decisions are made.