Release 1.13.1: Introducing the CLV Prediction ML Solution Template

We’re excited to share our third release for 2024, which not only includes a new ML Solution Template, but also a new database integration, and several improvement to other aspects of the Engine.

New Customer Lifetime Value Prediction Template

Our ML Solution Templates streamline the process of building a machine learning pipelines, and our newest template is a highly-requested use case; Customer Lifetime Value Prediction.

With the template, you can predict the total transaction amount, or the total number of transactions for each of your customers, allowing you to quickly identify your highest value customers quickly and easily.

Simply create an app from the Project Home page, and you’ll see the new Customer Lifetime Value option under our Machine Learning Solution Templates section.

We’ve also added a new CLV prediction application in the playground project loaded into your account, so you can check out the template straight away.

Select Customer Lifetime Value from project home page

Select Use this template to create the application CLV prediction

New business-friendly features for templates

Speaking of templates, we’re always looking to improve our existing ones, so we’ve updated and added some features to all templates.

Updated App Summary page

Firstly, we’ve updated the App Summary page for templates, so that they now clearly display the business requirements and training data that have been used for the application.

App Summary page of a Customer Lifetime Value Prediction app

Historical data insights

Secondly we’ve added a feature allowing you to analyse training data, providing insights about the historical data, relating to the template being used.

Analysis charts of historical CLV

Prediction Explanation

Finally, we’ve added the ability to view the details for a customer in the prediction explanation feature, as well as updating to a cleaner look suited for business users.

Customer details window transactional churn-png

Detailed Model Training Status

Our next improvement for the release relates to the App Summary page for apps built with both FlexiBuild Studio and ML Solution Templates. The model training process has been made clearer, showing in more detail the progress status including queued, evaluating, training and ready.

When multiple models are training, the best performing model that is ready at that moment now is displayed, so you can decide whether it is enough for you to start making predictions, or if you need to wait for other models to finish training.

New model training statuses shown in the model listing page

New Database Integration - Single Store

Integrations are important to make the process of importing and exporting data to and from the Engine a seamless experience, and in this release we’ve added a new integration to the SingleStore database.

AIAE Integrations 1.13.1

Re-watching the quick start video

Our initial quick start video helps you to get started with the Engine, so we’ve added the ability to rewatch it from the help menu. Whether you skipped it or just want a refresher, the video will help you better understand the capabilities of the Engine


Wrapping Up

So that wraps up the 1.13.1 release! We’re always looking to add new features and improve current ones, and there will be plenty more releases as we continue through Q2 2024.

Similar posts