AI in the cloud: AWS makes machine learning more accessible for developers
Amazon Web Services Inc.’s re:Invent conference is still nearly a week away, but you wouldn’t know it from the sheer number of new products and updates its announced in recent days — especially in artificial intelligence, likely to be a key focus of the conference.
Following last week’s storage announcements and its “internet of things” updates on Monday, AWS today introduced new features aimed at making it easier for developers to add AI predictions to their applications and services.
The central idea is to put Amazon’s machine learning technology in reach of more developers, AWS principal Matt Asay said in a blog post. Machine learning predictions will soon be able to run on unstructured or relational data in Amazon S3, its main storage service, and Amazon Aurora, which is a cloud-hosted MySQL and PostgreSQL-compatible relational database service.
What that means is that customers will be able to train machine learning models in SQL using Aurora or AWS Athena, which is an interactive query service for analyzing data in S3.
The benefits will also extend to AWS QuickSight, which is the company’s data visualization tool for creating and publishing dashboards that highlight AI insights. With the updated features, QuickSight will be able to visualize and report model predictions from machine learning services such as AWS SageMaker, Asay said.
The new features are designed to reduce the amount of custom code that needs to be written, managed and supported in production. Previously, developers would need to copy data from stores and transform it into a compatible format and then feed it to their machine learning models, which not only takes time but also complicates governance and security, Asay said.
“[Now,] you don’t need to [make calls] from your application, making it simpler to add predictions to your applications without having to build custom integrations or learn separate tools,” Asay wrote. “Now anyone who can write SQL can make — and importantly, use — predictions in their applications without any custom code.”
In other AI-related news today, Amazon announced that it has snagged another major new customer: European satellite TV broadcaster ProSiebenSat.1 Media SE. The Germany-based company said it has chosen AWS as its “primary cloud provider” and plans to integrate Amazon’s machine learning services into every aspect of its business to automate processes and develop new, personalized products.
Separately, Amazon cloud rival Google LLC has provided an update on its own AI and machine learning efforts, highlighting how three major enterprises are using its AutoML Vision service to improve their manufacturing visual inspection processes.
The enterprises include the semiconductor firm GlobalFoundries, which is using AutoML Vision to build a visual inspection tool capable of detecting defects in its Wafer Map and Scanning Electron Microscope images. A second customer is Siemens Aktiengesellschaft, which is using Auto ML Vision to create a “Factory AI service” that enables it to build prototypes of new products more quickly.