What can I do with OneAI? What is included in the toolkit?

Answer: One AI capabilities are divided into three types: machine learning models, data augmentations, and embedded insights  

In the machine learning category, users can create classification (binary and multi-class) and regression models. One AI Recipes, a guided workflow for easily creating machine learning models, allow you to predict Voluntary Attrition, Involuntary Attrition, New Hire Failure, High Performance, or perform Promotions.  If you are interested in predicting other things, you can work with your CS lead on creating a custom or advanced model - as long as you have a metric defining what you would like to predict, you can build it! The results of these models can be integrated into your data model and modeled onto storyboards. The One AI team is always working on adding more model types to the recipes page, so please continue to read release notes and check back for more. 

Moving on to data augmentations, commute time and distance is available.  This data is obtained by passing zip code combinations from your data to a mapping service and returning the times and distances. You can also create a variety of the commute time augmentation - closest office commute time. The results tell users which office is closest (time or distance) to each employee and can be used to assign employee's to the "closest" office.  Data from these augmentations can be integrated into your data model and modeled onto storyboards.

Finally, embedded insights are statistical capability integrated directly into One Model storyboard tiles.  Current integrated services include trend forecasting, table insights, and scatterplot correlations with line of best fit. If you have created a tile that meets the necessary criteria for a specific embedded insight, a lightbulb icon will appear in the upper righthand corner of the tile on your storyboard. You can click the lightbulb to run the embedded insight. 

Table Insights embedded insight on a list tile in a storyboard

 

Was this article helpful?

0 out of 0 found this helpful

Comments

0 comments

Please sign in to leave a comment.