AWS Provides Six Key New Reasons to Adopt and Use Amazon SageMaker

The News: Amazon Web Services, Inc. (AWS), an Amazon.com, Inc. company, announced six new capabilities for its machine learning (ML) service, Amazon SageMaker, that targets making ML more accessible and cost effective. The announcements at AWS re:Invent bring together new capabilities, including a no-code environment for creating accurate machine learning predictions, more accurate data labeling using highly skilled annotators, a universal Amazon SageMaker Studio notebook experience for greater collaboration across domains, a compiler for machine learning training that can make code more efficient, automatic compute instance selection machine learning inference, and serverless compute for machine learning inference. Read the AWS Press Release here.

AWS Provides Six Key New Reasons to Adopt and Use Amazon SageMaker

Analyst Take: I am pleased to see AWS up its market outreach and ecosystem influence with the introduction of the additional Amazon SageMaker capabilities, especially across the ML and AI ecosystem. The new Amazon SageMaker additions consist of the following six capabilities:

• Amazon SageMaker Canvas no-code ML predictions: Amazon SageMaker Canvas is designed to expand access to ML by providing business analysts (i.e., line-of-business employees supporting finance, marketing, operations, and human resources teams) with a visual interface that allows them to potentially develop more accurate ML predictions on their own and all without requiring any ML experience or having to write a single line of code. Leveraging built-in integration with Amazon Forecast, I expect that the Canvas offering can ease the access, importing, and viewing of data from both cloud and on-premises data sources and is poised to improve data collaboration by sharing models with data scientists.
• Amazon SageMaker Ground Truth Plus expert data labeling: Amazon SageMaker Ground Truth Plus is a managed data labeling service that uses an expert workforce with built-in annotation workflows that aims to deliver data for training ML models faster and at lower cost with no coding required. I anticipate that this new offering can increase data quality through ML-powered data labeling, access to highly skilled data labelers, reduce data labeling costs with assistive labeling features, and make data labeling accessible to data operations and program managers.
• Amazon SageMaker Studio universal notebooks: A notebook for Amazon SageMaker Studio (a complete integrated development environment for ML) provides a single, integrated environment to perform data engineering, analytics, and machine learning. I foresee this new capability enabling the built-in development environment needed for writing, monitoring, and debugging interactive Hive and Spark queries in Amazon SageMaker Studio notebooks.
• Amazon SageMaker Training Compiler for ML models: Amazon SageMaker Training Compiler is a new ML model compiler that automatically optimizes code that seeks to use compute resources more effectively and reduce the time it takes to train models by up to 50%. I view the new Amazon SageMaker Training Compiler’s integration with the versions of TensorFlow and PyTorch in Amazon SageMaker that have been optimized to run more efficiently in the cloud, as ready to play a key role in enabling data scientists to increasingly use their preferred frameworks to train ML models through more efficient use of GPUs.
• Amazon SageMaker Inference Recommender automatic instance selection: Amazon SageMaker Inference Recommender helps customers automatically select the best compute instance and configuration (e.g., instance count, container parameters, and model optimizations) to power a particular ML model. I believe the new Inference Recommender is ready to ensure that customers can review benchmark results in Amazon SageMaker Studio and better examine the tradeoffs between different configuration settings including latency, throughput, cost, compute, and memory factors.
• Amazon SageMaker Serverless Inference for ML models: Amazon SageMaker Serverless Inference supports pay-as-you-go pricing inference for ML models deployed in production. Customers are looking to optimize costs when using ML, and I see this as becoming increasingly important for applications that have intermittent traffic patterns with long idle times.

AWS needed to announce these six new Amazon SageMaker capabilities to build on the market inroads and presence the SageMaker Data Wrangler, SageMaker Processing, SageMaker Feature Store, and SageMaker Clarify offerings have established in the structured data realm as well as the SageMaker Ground Truth and SageMaker Ground Truth Beacon offerings in the unstructured data realm. As a result, I see AWS rapidly broadening its ability to drive ecosystem-wide adoption of comprehensive ML capabilities for data preparation and annotation across structured data and unstructured data applications.

Why ML Technology is Playing a More Influential Role in Driving Business Outcomes and Innovation

Overall, I see ML becoming more integral in the strategic decision making of organizations since it enables wider access to large volumes and types of data, particularly in the development of actionable insights and underpinning the automation of business processes and operations. ML-enabled data processing and workload efficiencies can allow for more affordable data storage solutions and assure that data is more widely accessible and supportive of a wider array of applications. ML-powered platforms can deliver the completion of calculations faster that accelerate data compilation, aggregation, distribution, storage, and curation processes.

I see the new Amazon SageMaker capabilities further attesting to the rapidly expanding influence of ML and AI technology across the digital ecosystem. This week at AWS re:Invent, the company shared how Amazon SageMaker is already supporting 1M+ labeling tasks per day, how customers are running millions of models with billions of parameters, and making 100B+ predictions per month.
Of note, Amazon Sagemaker already includes tens of thousands of customers, including Airbnb, AstraZeneca, Aurora, BMW Group, Capital One, Cerner, Discovery, Hyundai, Intuit, Litterati, NFL, Provectus, Siemens Energy, Tyson, Vanguard, and VIZIO who use the service to train ML models of all sizes, providing additional validation to the growing influence and mind share of cloud-based ML services.

Key Takeaways on AWS Introducing Six New Amazon SageMaker Capabilities

Through the introduction of the six new Amazon SageMaker capabilities, I believe AWS can make substantial inroads and strides in expanding customer use of its ML offerings as well as onboard new users and adopters of SageMaker ML technology. The capabilities are well-suited to enable more organizations to better understand the competitive benefits of using ML technology by providing comprehensive data preparation capabilities and easing the path to learning ML. Now AWS rivals such as Azure, Google Cloud, Oracle, IBM/Red Hat, HPE GreenLake, and Alibaba, will need to refresh their ML propositions to counter the expanded capabilities of the Amazon SageMaker solution.

This article includes insight from Senior Analyst and Research Director, Ron Westfall.

Disclosure: Futurum Research is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.

Other insights from Futurum Research:

Amazon Provides Free Access to Legal Entity Identifier Dataset on the AWS Cloud to Help Businesses Protect Investments from Climate-change Related Risks

AWS Announced Three New Capabilities for Amazon Connect

Amazon Results Miss Expectations as Investments in Growth Continue

Image Credit: AWS

The original version of this article was first published on Futurum Research.