Build a predictive model using AI
Released on March 5, 2026, this update introduces a functional enhancement powered by artificial intelligence. It enables users to quickly generate predictive competency models while streamlining and accelerating the configuration process.
How to use AI to create a predictive model
From the Key Predict application menu, go to the Predictive Model tab, then click Add.
Please note that the process remains the same whether you create your predictive model from the dedicated menu or directly when setting up a campaign. Once on the model creation page, enter your HR project or the relevant job offer in the designated field.
When writing your description, make sure to follow these guidelines:
- The text must contain at least 20 characters.
- It should correspond to a job offer or an HR project.
- It should not be a general list of skills.
After completing your description, click Validate to launch the automatic model generation.
Once the analysis is complete, the model is automatically generated based on your input. It includes suggested roles, job functions, and competencies, along with an importance level assigned to each.
đ This automated generation significantly speeds up the model creation process while providing relevant suggestions tailored to your needs.
At this stage, the model remains fully customizable: you can adjust the suggested importance levels and add new competencies to refine the model and align it with your specific objectives.
The system automatically identifies the most relevant combinations of roles and functions based on your description. These elements can be easily reviewed and adjusted by clicking on Edit.
If you wish to modify these suggestions, you can disable Key Predictâs smart assistance to display the detected roles and functions, and then adjust them as needed.
Once your changes are made, click Validate to regenerate the model with the updated parameters.