Praexia: Scorecard and Credit Modelling Toolkit


Praexia: Scorecard and Credit Modelling Toolkit


Empowers decision scientists with a complete credit modelling lifecycle for large and small datasets.

About the offer:

Praexia is a suite of scorecard modelling libraries specifically aimed at credit risk modelling. Get started with your first model using the example notebooks or make your own model using the tools in the libraries. The service operates within the Jupyter environment, where JupyterHub aids in security and access control and JupyterLab is used to create and manage your notebooks.

Intended users:

• Scorecard modellers
• Credit risk analysts and modellers
• Data scientists

What is included:

Example notebook
This notebook illustrates the possibilities of the libraries. Use the notebook as-is, or modify it, to create your first scorecard model.
Access to the libraries
Want more flexibility and customisation in your scorecard model? The libraries are readily available for you to build a scorecard that fits your exact needs.
JupyterHub ensures the security of your valuable data and manages who has access to it.

Key features:

• Automatic binning
• Manual binning adjustments
• Weight of Evidence transformations
• Plots and interactive widgets
• Reject inference
• Variable selection
• Performance and stability checking
• Monitoring

What our customers say

“Praexia allows us to improve on all aspects of risk management by providing a single toolkit for data analysis and preparation, modelling, deployment and monitoring. It allows us to use the latest tools and techniques, without sacrificing the transparency, robustness, customisation and efficiency we expect.” - Credit Modeller, Matogen AI (Financial Services Industry: 2024).

Startup instructions:

• Start the VM
• The web-app runs on port 8889
• Copy the following into the browser search bar "_vm ip_:8889"
• The login details will be displayed in the Boot Diagnostics for the first 24 hours
• After logging in, it is recommended that you change the login password from a new terminal in the lab environment on the VM
• Open a new terminal in Jupyter Lab and run the following three lines
• source activate praexia-env
• ipython kernel install --user --name=praexia-env
• ccc init pd

The setup is now complete. You can now either go to the example notebook or start creating your own notebook. Simply select the praexia-env kernel in your notebook, and you'll be on your way! Note that only a single user can use a single instance.