AWS Adds AI Agent to SageMaker for Model Customization
SEATTLE — Amazon Web Services has introduced an agentic workflow capability in its SageMaker AI platform that allows developers to customize machine learning models using natural language instructions, the company announced in a blog post.
The new feature deploys an AI coding agent that handles the full model customization lifecycle — from use case definition and data preparation through technique selection, evaluation, and deployment — based on plain-language descriptions from developers, according to the AWS Machine Learning Blog.
Rather than requiring ML engineers to manually navigate each step of the fine-tuning process, the agent-guided system uses what AWS calls “agent skills” to streamline the workflow. Developers describe their customization needs conversationally, and the agent orchestrates the technical implementation.
AWS said the agentic approach could reduce the time and specialized knowledge required to move from a base model to a production-ready customized version, targeting enterprise teams without deep MLOps expertise.
