Release 1.14.0, Aug 2024

August 2024 marks the AI & Analytics Engine's 1.14.0 release, introducing the newest ML Solution Template, Demand Forecasting.

We are pleased to announce the following new features and improvements to the AI & Analytics Engine, in our release for Q3 2024.

Demand Forecasting Template 
Redesign of Error and Notification Messaging 

Demand Forecasting Template

Our newest ML Solution Template will allow you to build an application to forecast demand for your product in a retail setting. Using this template, you can predict how much of a certain product you will need to fulfil demand over a given time period, with an added option of showing confidence intervals. The template brings:

  • Flexibility to define business requirements

  • Ability to forecast at the finest granularity levels as well as higher aggregate levels with a single application

  • Automation of continuous forecasting, evaluation and retraining.

  • Options to integrate your data source for automatic updates and directing output data

You’ll be able to access the new template via the Build new app section of the AI & Analytics Engine. Follow the complete walkthrough.

Demand Forecasting - Sample OutputSample output of the Demand Forecasting Template

 

Demand Forecasting Continuous Evaluation and ForecastingContinuous Evaluation and Forecasting in the Demand Forecasting Template

 

Demand Forecasting Model LeaderboardModel Leaderboard of a Demand Forecasting Application

Redesign of Error and Notification Messaging

Across our platform, to improve usability, we have made some changes to our error and notification messaging. These changes include:

  • Improvements to notification error messages to provide more context and details for exceptions.

  • Added new notifications to a wider range of system errors.

  • Added support for third party API errors in the platform.

  • Fixed existing errors in the notification system.

These new error and notification errors will allow you to better understand where and how to move forward in the case of system errors to ensure you can resume building, improving and deploying your models.