Introducing AutoAI for Watson Studio

We are delighted to announce the general availability of AutoAI in Watson Studio . We first previewed this in February for Think! 2019. With this release, we extend the portfolio of best-in-class data science and machine learning tools available in Watson Studio, the premier enterprise data science platform. This release focuses on the automation of the end-to-end model building lifecycle, from data ingestion to model deployment. With this release, Watson Studio continues its mission of boosting the productivity of data science teams large and small.

AutoAI in Watson Studio automates tasks that typically take data scientists days or weeks.

See it in action:

IBM’s Digital Technical Engagement team introducing AutoAI Experiments

To get started, any Watson Studio user can now add an AutoAI Experiment asset to a new or existing project. Once added, simply select a data set and a field to predict or classify and the automation handles the rest.

View stages of optimization in real time during training

Once the AutoAI training job completes, you can easily compare the machine learning pipelines to evaluate performance and gain intuition about the feature importance of fields used for the model from the raw data. After evaluating, any of the pipelines can easily be saved to your Watson Machine Learning repository and deployed as a web service deployment with a couple clicks.

View candidate pipelines in real time during AutoAI training

Benefits of using AutoAI for individuals and teams solving problems with AI

Related to this announcement, we will also be phasing out the Watson Machine Learning model builder tool on July 31, 2019. Models trained with Model Builder and deployed to Watson Machine Learning will continue to be supported, but following this date, no new models can be trained using Model Builder. We encourage all Model Builder users to begin leveraging AutoAI Experiments for their automated model building use cases.

Learn more:

• Update! Register for our upcoming webinar on automating feature engineering
• Join our webinar to go deeper on all things AutoAI with Julianna Delua, Ruchir Puri, and me!
• IBM Watson Studio AutoAI
• Create your Watson Studio account
• For more detailed guidance on using AutoAI Experiments in Watson Studio see our documentation or follow along with a tutorial.