Do not input private or sensitive data. View Qlik Privacy & Cookie Policy.
Skip to main content

Announcements
Join us in NYC Sept 4th for Qlik's AI Reality Tour! Register Now
cancel
Showing results for 
Search instead for 
Did you mean: 
Ernests
Partner - Creator
Partner - Creator

AutoML possibilities in Qlik Sense Enterprise SaaS

Hi Experts, 

I have some trouble understanding what I can do with autoML. I am using Qlik Sense Enterprise SaaS and the question is very general.

The question. Is it possible to create deployed model which regularly takes data and predict churn possibility so I can use the results in Qlik Sense app as dataset? Or the only option for my license is to train prediction model and manually upload data which need to be evaluated every time?

As I understood from link below, this tier is not suitable for production use cases. For more comprehensive capabilities, I have to consider a paid tier of Qlik AutoML.

Introducing automated machine learning | Qlik Cloud Help

Thanks!

Labels (3)
1 Solution

Accepted Solutions
michalnurzynski
Partner - Contributor II
Partner - Contributor II

Hi
Yes, you get it correctly. To be more precise, without a paid AutoML license you can deploy two models trained with some limitations and you'll need to run the predictions manually each time. 
In fact, since we have model approval, you can have two "active" models but a lot more models created but not activated.  
These models can be used for production deployment but limitations in their training may limit their precision, so their value depends on the use-case.
Regards

View solution in original post

1 Reply
michalnurzynski
Partner - Contributor II
Partner - Contributor II

Hi
Yes, you get it correctly. To be more precise, without a paid AutoML license you can deploy two models trained with some limitations and you'll need to run the predictions manually each time. 
In fact, since we have model approval, you can have two "active" models but a lot more models created but not activated.  
These models can be used for production deployment but limitations in their training may limit their precision, so their value depends on the use-case.
Regards