Amazon Web Services (AWS) announced new capabilities in Amazon Bedrock and Amazon SageMaker AI at its re:Invent conference on Wednesday, aiming to simplify the creation and fine-tuning of custom large language models (LLMs) for enterprise customers. These additions follow the recent introduction of Nova Forge, a service for training custom Nova AI models.
The cloud provider is introducing serverless model customization within SageMaker AI, a feature designed to enable developers to commence model building without the need to manage compute resources or underlying infrastructure. Ankur Mehrotra, general manager of AI platforms at AWS, stated in an interview that developers can access these serverless capabilities through either a self-guided point-and-click interface or an agent-led experience that responds to natural language prompts. The agent-led feature is currently launching in preview. According to Mehrotra, this capability allows a customer in the healthcare sector, for example, to direct SageMaker AI with labeled data to fine-tune a model for understanding specific medical terminology. The service supports customization for Amazon’s proprietary Nova models and selected open-source models, including DeepSeek and Meta's Llama.
Concurrently, AWS is launching Reinforcement Fine-Tuning in Bedrock. This feature allows developers to specify a reward function or select a pre-set workflow, enabling Bedrock to automate the model customization process from initiation to completion.
Frontier LLMs, which represent the most advanced AI models, and their customization are a central focus for AWS at this year's conference. Mehrotra noted that many customers seek methods to differentiate their solutions if competitors have access to the same foundational models. He identified the ability to create customized models, optimized for specific brands, data, and use cases, as key to solving this challenge.
A July survey by Menlo Ventures indicated that enterprises have shown a preference for AI models from Anthropic, OpenAI, and Gemini. AWS’s expanded focus on custom and fine-tuned LLMs through these new services is positioned to enhance its competitive standing in the enterprise AI market.
