Skip to content

Usage

Deployed Resources

Once deployed, you should see the following in your AWS account:

Naming convention: <org>-<env>-<domain>-<module>-<resource>

SSM parameters: /<org>/<domain>/<module>/<resource-type>/<resource-name>/<attribute>

Resource Deployed Name / SSM Path Config Reference
Audit S3 Bucket <org>-dev-shared-audit-audit
/<org>/shared/audit/bucket/name		 deployed by shared/audit.yaml
Audit KMS Key <org>-dev-shared-audit-audit
/<org>/shared/audit/kms/arn		 deployed by shared/audit.yaml
CloudTrail Trail <org>-dev-shared-audit-trail-s3-audit trails.s3-audit in shared/audit-trail.yaml
SMUS Domain <org>-dev-sagemaker-domain-domain1
/<org>/sagemaker/domain/domain/domain1/name		 domains.domain1 in sagemaker/domain.yaml
SMUS Project Profile <org>-dev-sagemaker-projects
/<org>/sagemaker/projects/project/name		 configured in sagemaker/projects.yaml
IAM Role <org>-dev-shared-roles-data-admin
/<org>/shared/generated-role/data-admin/id		 generateRoles.data-admin in shared/roles.yaml
Glue Catalog Encryption Account-level setting deployed by glue-catalog module
(not configurable)
LakeFormation Data Lake Settings Account-level setting deployed by shared/lakeformation-settings.yaml
(not configurable)

Post-Deployment Steps

1. Access the SMUS Portal

  1. Assume the <org>-dev-shared-roles-data-admin role in your deployment account.
  2. Navigate to SageMaker Unified Studio in the AWS Console.
  3. Select the deployed domain (<org>-dev-sagemaker-domain-domain1).
  4. Click Open portal.

2. Verify SSO Access

  1. SSO users in the team1 and team2 groups should be able to log into the portal.
  2. If users cannot access the portal, verify their SSO group IDs match the values configured in mdaa.yaml context (team1-group-sso-id, team2-group-sso-id).

3. Create a Project

  1. In the SMUS portal, navigate to Projects.
  2. Select a project profile to create a new team workspace.
  3. Team members in the assigned SSO group will have access to the project's notebooks, SQL queries, and data catalog.

4. Start Working

Available capabilities in the portal: - JupyterLab notebooks — ML experimentation and data analysis - SQL queries — Query data in Glue databases via Athena - Data catalog — Discover and subscribe to data assets - MLflow — Track experiments and model versions

Scaling Up

For more complex organizations with teams operating across multiple accounts, see the SMUS Data Mesh starter kit.