It's been nearly a decade since Amazon Web Services (AWS), Amazon's cloud computing division, announced SageMaker, its platform for constructing, training, and deploying AI models. While AWS has focused on significantly expanding SageMaker's capabilities in recent times, this yr the goal was streamlining.
At its re:Invent 2024 conference, AWS introduced SageMaker Unified Studio, a central place to seek out and work with data from across the enterprise. SageMaker Unified Studio brings together tools from other AWS services, including the prevailing SageMaker Studio, to assist customers discover, prepare and process data to construct models.
“We are seeing a convergence of analytics and AI, with customers using data in increasingly connected ways,” said Swami Sivasubramanian, vice chairman of information and AI at AWS, in a press release. “The next generation of SageMaker brings together capabilities to provide customers all of the tools they should process data, develop and train machine learning and generative AI models, right in SageMaker.”
With SageMaker Unified Studio, customers can publish and share data, models, apps, and other artifacts with members of their team or a bigger organization. The service offers data security controls and customizable permissions, in addition to integrations with AWS's Bedrock model development platform.
AI is integrated into SageMaker Unified Studio – more specifically, Q Developer, Amazon's programming chatbot. In SageMaker Unified Studio, Q Developer can answer questions like “What data should I exploit to get a greater idea of ​​product sales?” or “Generate SQL to calculate total sales by product category.”
AWS explained in a blog post: “Q Developer (can) support development tasks similar to data discovery, coding, SQL generation and data integration” in SageMaker Unified Studio.
Beyond SageMaker Unified Studio, AWS has launched two small additions to its SageMaker family of products: SageMaker Catalog and SageMaker Lakehouse.
SageMaker Catalog allows administrators to define and implement access policies for AI apps, models, tools, and data in SageMaker using a single permissions model with granular controls. Meanwhile, SageMaker Lakehouse provides connections from SageMaker and other tools to data stored in AWS data lakes, data warehouses, and enterprise applications.
According to AWS, SageMaker Lakehouse works with all tools compatible with the Apache Iceberg standards – Apache Iceberg is the open source format for big evaluation tables. If desired, administrators can apply access controls to all data across all analytics and AI tools that SageMaker Lakehouse uses.
In a somewhat related development, SageMaker is alleged to now work higher with software-as-a-service applications because of latest integrations. SageMaker customers can access data from apps like Zendesk and SAP without having to first extract, transform, and cargo that data.
“Customers can have data spread across multiple data lakes in addition to an information warehouse and would profit from a simple strategy to unify all of this data,” AWS wrote. “Now customers can use their favorite analytics and machine learning tools on their data, no matter how and where it’s physically stored, to support use cases similar to SQL analytics, ad hoc querying, data science, machine learning and generative AI. ”
