AWS intros SageMaker Canvas no-code machine learning service

AWS now has a no-code tool that enables enterprises to predict business outcomes based on data without needing machine learning expertise.

The tech giant introduced SageMaker Canvas at its AWS re:Invent conference on Nov. 30. The tool — an addition to the SageMaker suite of AI services — automatically cleans and combines an organization’s data and can create hundreds of models behind the scenes, select the best performing one, and generate new individual or batch predictions, according to AWS.

SageMaker Canvas uses a visual point and click interface, enabling users to make predictions without needing machine learning experience or writing code.

What SageMaker Canvas offers

“Canvas uses terminology and visualization that are already familiar to analysts and complements the data analysis tools they’re already using,” Adam Selipsky, AWS CEO, said during his keynote presentation.

With SageMaker Canvas, users can browse and access petabytes of data from both cloud and on-premises data sources such as Amazon s3, Redshift and local files. Once Canvas creates predictive models, users can publish the results, plan and interpret models to share dashboards and collaborate with other data analysts.

The problem SageMaker Canvas addresses

The idea is to automate the application development process, said Sid Nag, an analyst at Gartner.

“With the Canvas offering, [AWS is] extending the SageMaker functionality to make the life of the developer easier,” Nag said.

He added that AWS is giving users the tools that enable  them to use a visual process to generate the code for machine learning applications that otherwise would have needed to be written by a developer or data scientist.

Nag said that since machine learning applications can be detailed and complex, the SageMaker Canvas system will help simplify it for users who want to developer applications with AI.

“Anything that simplifies the generation of application that leverage machine learning is a good thing,” he said.

Nag added that the technology will also give enterprises that are already writing machine learning applications on AWS more options and drive more adoption and more application development on AWS.

As to the possible challenges that enterprises using the new tool might face or how it compares to other no-code products, Nag said that users will have to wait and see.

According to AWS, SageMaker Canvas supports multiple problem types including binary and multi-class classification, numerical regression and time series forecasting – so users can create applications to do fraud detection and inventory optimization, among other things.

Canvas is AWS’s latest capability in the SageMaker line, which also includes SageMaker Studio, SageMaker Experiments, SageMaker Debugger and others.

SageMaker Canvas is generally available in the United States and Europe now.