Access subscribed data in SageMaker
Query the datasets you've subscribed to from inside SageMaker Studio. Approved subscriptions are provisioned to your project's Data Lake automatically, ready to query through Amazon Athena.
Before you start
- You are onboarded to the MedTech Connect MLOps platform
- You have a project on the MedTech Connect MLOps platform
- You have subscribed to at least one dataset and the Data Asset Owner has approved it (see "Subscribe to data using MedTech Connect")
Steps
- 01
Open SageMaker Studio for your project
Launch SageMaker Studio from the MLOps platform (see the "Access SageMaker Studio" guide). Studio opens scoped to your project, with your approved subscriptions already provisioned to the project environment.
- 02
Confirm the subscription is approved
In your project, open the "Data" tab and select "Requested data" to confirm the dataset shows as approved. Approved managed assets are added to your project's Data Lake automatically, and only assets approved for your project are visible.
- 03
Open a notebook or query editor
In Studio, open a JupyterLab notebook (or the SQL query editor) in your project environment to start working with the data.
- 04
Query the data through Athena
Subscribed managed assets are queryable through Amazon Athena in your project's database. Use the Athena deep link from the asset's governance page, or query from a notebook with a SQL client (such as awswrangler or PyAthena) pointed at your project's Athena workgroup and the subscribed table.
- 05
Analyze or train on the results
Load query results into a DataFrame for analysis, visualization, or model training. Access stays governed end to end, and usage against the subscribed asset is logged.