October 3, 2023
Voiced by Polly

Amazon SageMaker Canvas gives enterprise analysts with a visible interface to unravel enterprise issues utilizing machine studying (ML) with out writing a single line of code. Since we launched SageMaker Canvas in 2021, many customers have requested us for an enhanced, seamless collaboration expertise that allows information scientists to share skilled fashions with their enterprise analysts with a number of easy clicks.

In the present day, I’m excited to announce that you would be able to now convey ML fashions constructed anyplace into SageMaker Canvas and generate predictions.

New – Deliver Your Personal Mannequin into SageMaker Canvas
As an information scientist or ML practitioner, now you can seamlessly share fashions constructed anyplace, inside or outdoors Amazon SageMaker, with your enterprise groups. This removes the heavy lifting on your engineering groups to construct a separate software or consumer interface to share ML fashions and collaborate between the completely different elements of your group. As a enterprise analyst, now you can leverage ML fashions shared by your information scientists inside minutes to generate predictions.

Let me present you ways this works in observe!

On this instance, I share an ML mannequin that has been skilled to determine clients which are doubtlessly vulnerable to churning with my advertising analyst. First, I register the mannequin within the SageMaker mannequin registry. SageMaker mannequin registry enables you to catalog fashions and handle mannequin variations. I create a mannequin group known as 2022-customer-churn-model-group after which choose Create mannequin model to register my mannequin.

Amazon SageMaker Model Registry

To register your mannequin, present the situation of the inference picture in Amazon ECR, in addition to the situation of your mannequin.tar.gz file in Amazon S3. You can even add mannequin endpoint suggestions and extra mannequin data. When you’ve registered your mannequin, choose the mannequin model and choose Share.

Amazon SageMaker Studio - Share models from model registry with SageMaker Canvas users

Now you can select the SageMaker Canvas consumer profile(s) throughout the identical SageMaker area you need to share your mannequin with. Then, present further mannequin particulars, corresponding to details about coaching and validation datasets, the ML downside sort, and mannequin output data. You can even add a be aware for the SageMaker Canvas customers you share the mannequin with.

Amazon SageMaker Studio - Share a model from Model Registry with SageMaker Canvas users

Equally, now you can additionally share fashions skilled in SageMaker Autopilot and fashions obtainable in SageMaker JumpStart with SageMaker Canvas customers.

The enterprise analysts will obtain an in-app notification in SageMaker Canvas {that a} mannequin has been shared with them, together with any notes you added.

Amazon SageMaker Canvas - Received model from SageMaker Studio

My advertising analyst can now open, analyze, and begin utilizing the mannequin to generate ML predictions in SageMaker Canvas.

Amazon SageMaker Canvas - Imported model from SageMaker Studio

Choose Batch prediction to generate ML predictions for a complete dataset or Single prediction to create predictions for a single enter. You may obtain the leads to a .csv file.

Amazon SageMaker Canvas - Generate Predictions

New – Improved Mannequin Sharing and Collaboration from SageMaker Canvas with SageMaker Studio Customers
We additionally improved the sharing and collaboration capabilities from SageMaker Canvas with information science and ML groups. As a enterprise analyst, now you can choose which SageMaker Studio consumer profile(s) you need to share your standard-build fashions with.

Your information scientists or ML practitioners will obtain an identical in-app notification in SageMaker Studio as soon as a mannequin has been shared with them, together with any notes from you. Along with simply reviewing the mannequin, SageMaker Studio customers can now additionally, if wanted, replace the info transformations in SageMaker Knowledge Wrangler, retrain the mannequin in SageMaker Autopilot, and share again the up to date mannequin. SageMaker Studio customers may also advocate an alternate mannequin from the record of fashions in SageMaker Autopilot.

As soon as SageMaker Studio customers share again a mannequin, you obtain one other notification in SageMaker Canvas that an up to date mannequin has been shared again with you. This collaboration between enterprise analysts and information scientists will assist democratize ML throughout organizations by bringing transparency to automated selections, constructing belief, and accelerating ML deployments.

Now Obtainable
The improved, seamless collaboration capabilities for Amazon SageMaker Canvas, together with the power to convey your ML fashions constructed anyplace, can be found in the present day in all AWS Areas the place SageMaker Canvas is on the market with no modifications to the present SageMaker Canvas pricing.

Begin collaborating and produce your ML mannequin to Amazon SageMaker Canvas in the present day!

— Antje