AWS Boosts SageMaker for Generative AI

At AWS re:Invent, Amazon Web Services (AWS), a subsidiary of Amazon.com, Inc. (NASDAQ: AMZN), announced four groundbreaking updates to Amazon SageMaker AI. These advancements aim to help users start faster with popular publicly available models, enhance training efficiency, reduce costs, and offer flexibility to accelerate generative AI model development. Amazon SageMaker AI, an end-to-end service trusted by hundreds of thousands of customers, simplifies the process of building, training, and deploying AI models with fully managed infrastructure, tools, and workflows.

One of the key updates is the expansion of Amazon SageMaker HyperPod, which now includes three new features designed to improve model training. Customers can use curated model training recipes for popular models like Llama 3.2 90B and Mistral 8x22B, enabling faster customization for specific use cases. Flexible training plans allow users to set timelines and budgets, with HyperPod automatically managing compute resources to optimize efficiency. Enhanced task governance ensures compute accelerators are fully utilized and prioritized for critical tasks, helping reduce costs by up to 40%.

AWS has also introduced seamless integration with partner applications within Amazon SageMaker, enabling customers to easily discover and deploy generative AI tools from leading providers like Comet, Deepchecks, Fiddler AI, and Lakera. These applications support specialized tasks such as model performance monitoring and experiment tracking, with AWS Identity and Access Management (IAM) ensuring secure and compliant use. This integration reduces onboarding time for partner applications from months to weeks, giving customers greater flexibility and control over their AI development workflows.

Organizations such as Salesforce, Thomson Reuters, Fidelity, Luma AI, and Articul8 are already leveraging these new SageMaker capabilities to accelerate generative AI development. Startups and enterprises alike are using the enhanced features to train and fine-tune publicly available models, transforming their business operations with generative AI.

Dr. Baskar Sridharan, Vice President of AI/ML Services and Infrastructure at AWS, highlighted the continued innovation of SageMaker, which has introduced over 140 new capabilities since 2023. “With today’s announcements, we’re offering customers the most performant and cost-efficient model development infrastructure possible to help them accelerate the pace at which they deploy generative AI workloads into production,” he said.

With these latest updates, AWS strengthens its position as a leader in AI and machine learning, empowering organizations to scale generative AI applications faster and more cost-effectively. SageMaker continues to enable businesses like Intuit, Perplexity, and Rocket Mortgage to build foundation models and unlock the full potential of AI-driven innovation.