Qlik AutoML
No-code machine learning for accessible, powerful predictive analytics.. [Free Trial]
Last verified:
What is Qlik AutoML?
Qlik Predict (formerly Qlik AutoML) is a no-code machine learning tool that empowers analytics teams and business analysts to build, train, and deploy predictive models without requiring data science expertise or coding skills. The tool自动化ly finds patterns in historical data to create machine learning models that predict future outcomes like churn, revenue, or trends over time. It supports classification, regression, and multivariate time series forecasting with automatic seasonality detection.
Key features include a guided no-code workflow for model creation, SHAP-based explainable AI visuals that show what drives each prediction, full model lifecycle governance with automated documentation, continuous learning capabilities where models adapt to new data, and seamless integration with Qlik Cloud Analytics for embedding predictions directly into dashboards. Users can create batch predictions, real-time predictions via API, set alerts based on predictive insights, and run what-if scenarios with real-time calculations.
Qlik Predict is designed for business analysts, analytics teams, and organizations that want predictive analytics without hiring data scientists. It serves companies across industries for use cases like customer churn prediction, revenue forecasting, appointment cancellation reduction, automated risk scoring, and lead qualification. The tool is part of Qlik Cloud Analytics Premium and Enterprise tiers, serving over 40,000 customers including 75% of Fortune 500 companies.
Qlik AutoML pricing
Pricing model: Freemium
Qlik Predict is included in Qlik Cloud Analytics Premium and Enterprise tiers. Premium costs $2,750/month for 50 GB data (billed annually), includes predictive analytics powered by automated machine learning, additional GenAI capacity, anonymous access, and 10 GB max app size. Enterprise starts at 250 GB with custom quote, adds scale for large deployments, 15 GB apps standard, support for up to 40 GB apps (up to 50 GB with purchase), 3 GB Personal Space, multi-region tenants, and personalized customer success. Starter ($300/mo for 10 users, 10 GB) and Standard ($825/mo for 25 GB) do not include AutoML. Pricing uses capacity-based model with fixed annual fees. Additional data capacity purchased in 25 GB or 250 GB packs.
Qlik AutoML pros
- No-code interface requires no data science skills or programming
- SHAP-based explainable AI shows exactly what drives each prediction
- Native multivariate time series forecasting with automatic seasonality detection
- Full model lifecycle tracking and auditability for AI governance compliance
- Models adapt continuously to new data and user feedback without manual retraining
- Seamless integration into Qlik Sense dashboards and apps for operational predictions
- Unlimited experimentation included with Enterprise SaaS subscription
- Supports classification, regression, and time series in one interface
- Interactive SHAP visuals update dynamically with associative selections
- Batch predictions and real-time API predictions both supported
- What-if scenario testing with real-time calculations in apps
- Alerts and automations trigger based on predictive insights
- Works with diverse data sources via Qlik Data Flow integration
- Builds accurate models 80% faster than traditional methods
- Cloud-agnostic deployment with region selection options
- Bias detection supported for responsible AI practices
- Associative engine enables feature engineering without coding
Qlik AutoML cons
- Only available in Premium and Enterprise tiers, not in Starter or Standard plans
- Maximum 500 columns allowed in training datasets limits complex models
- API rate limit of 300 requests per minute constrains high-volume real-time use
- Not supported in Qlik Cloud Government environment
- Training fails if dataset exceeds subscription cell count limits
- Time series experiments only supported for datasets up to 1 GiB
- Bias detection only available for datasets under 1 GiB
- Qlik Predict analytics connector does not support time series models
- Limited customization compared to code-based ML platforms for advanced users
- Multi-sheet Excel files only import the first table
- Large datasets over 1 GiB are partially profiled causing potential training errors
- Real-time prediction APIs require add-on tiers beyond included subscription capacity
Frequently asked questions about Qlik AutoML
What is Qlik Predict and what does it do?
Qlik Predict is no-code machine learning software that finds patterns in historical data to make predictions on future data. It automates machine learning for analytics teams, enabling them to create ML experiments, generate models, and make predictions without writing code. Users can classify outcomes like churn, predict values like revenue, or forecast trends over time with full explainability using SHAP visuals.
Do I need data science skills to use Qlik Predict?
No. Qlik Predict features a simple code-free interface with guided workflows that handle complexity automatically. The tool auto-selects the best model approach based on your data, making machine learning accessible to business analysts and analytics teams without requiring a Ph.D. in data science or programming skills.
What subscription tiers include Qlik Predict?
Qlik Predict is included in Qlik Cloud Analytics Premium ($2,750/mo for 50 GB) and Enterprise (custom quote starting at 250 GB) tiers. It is not available in Starter ($300/mo) or Standard ($825/mo) plans. Premium adds predictive analytics powered by automated machine learning beyond Standard capabilities.
What types of machine learning models can I create?
Qlik Predict supports three model types: classification (for outcomes like churn), regression (for values like revenue), and time series forecasting (for trends over time). The tool automatically detects seasonality in time series data and fine-tunes algorithms for accuracy through guided workflows.
How does Qlik Predict explain its predictions?
Qlik Predict uses SHAP-based visualizations that show exactly what drives each prediction. These interactive visuals update dynamically as users explore data with associative selections, providing transparency into key factors influencing outcomes. This explainable AI helps build trust, meet AI governance standards, and take smarter action.
What are the dataset size limits for Qlik Predict?
Training datasets have maximum limits by type: CSV/Parquet/QVD support up to 2 GiB with 500 million cells and 500 columns; Others support up to 1 GiB with 100 million cells. Time series experiments and bias detection only work on datasets up to 1 GiB. Maximum 500 columns can be used as features in any experiment.
Can I deploy predictions in real-time?
Yes, Qlik Predict supports real-time predictions using the Machine Learning API for one or a handful of rows. Real-time prediction APIs are available with add-on tiers of Qlik Predict. You can also create batch predictions for datasets and use the Qlik Predict analytics connector to make predictions directly in applications.
Is Qlik Predict available in Qlik Cloud Government?
No. Qlik Cloud Government does not support Qlik Predict. This limitation applies to both the core AutoML functionality and the Qlik Predict analytics connector, which is also not supported in the Government environment.
How does Qlik Predict handle model updates and retraining?
Qlik Predict features continuous learning where models adapt to changing inputs, user feedback, and new data. This reduces the need for manual retraining or reengineering. Models automatically update as they encounter new data, keeping predictions current without requiring constant manual intervention.
What is the API rate limit for Qlik Predict?
Qlik Predict has an API rate limit of 300 requests per minute for real-time predictions. The Qlik Predict analytics connector is limited to 200k rows per request, sent in batches of 2k rows. For processing more rows, use a Loop within the data load script to process in batches.