[{"data":1,"prerenderedAt":12},["ShallowReactive",2],{"stack-tr-en-amazon-sagemaker":3},{"tagline":4,"description":5,"strengths":6,"useCase":11},"AWS's managed ML platform, from Studio notebooks to production endpoints.","SageMaker AI, the ML side of Amazon SageMaker, chains Studio (notebooks, JupyterLab, Code Editor), training and tuning jobs on managed instances, SageMaker Pipelines to orchestrate the steps, the Model Registry, and then real-time, serverless, asynchronous or batch endpoints watched by Model Monitor. JumpStart provides foundation and open models ready to fine-tune, and SageMaker Unified Studio ties it all to the data (S3, Redshift, Glue) and to Bedrock models. It's our choice when the client's infrastructure and data already live on AWS.",[7,8,9,10],"Studio: notebooks, JupyterLab and Code Editor on on-demand managed instances.","Pipelines and Model Registry: versioned steps, approval gates before deployment.","Real-time, serverless, asynchronous or batch endpoints, with Model Monitor.","JumpStart and Bedrock adjacency: foundation models to fine-tune or consume.","ML on AWS data, real-time scoring, demand forecasting, fine-tuning open models.",1792675980146]