Trusted Identity Propagation with Studio

Documentation for Amazon SageMaker AI's Trusted Identity Propagation feature—enabling enterprise identity management across AWS services with fine-grained access control and comprehensive auditing capabilities.

Role: Primary Author
Audience: Enterprise Administrators, Security Engineers, IAM Identity Center Administrators
Type: Architecture Documentation, Security Configuration, Integration Guides
Impact: Major Enterprise Feature, Cross-Service Integration

View the Live Documentation

Note on Live Documentation

As with all actively maintained documentation, these pages may have been updated by other contributors since original publication. The core structure and approach I established remain the foundation of these documentation nodes.

The Challenge

Trusted Identity Propagation addresses a critical enterprise need: maintaining user identity context as data scientists work across multiple AWS services. Before this feature, enterprises faced significant challenges:

  • Identity fragmentation: Users had to manage multiple IAM roles when accessing different AWS services, losing their individual identity context
  • Audit complexity: Tracking which specific user performed actions across services was difficult when all actions appeared under shared execution roles
  • Access control limitations: Fine-grained permissions based on user attributes (like group membership) were hard to implement across service boundaries
  • Cross-team coordination: Setting up the feature required collaboration between SageMaker administrators and IAM Identity Center administrators

My Approach

  • Multi-audience documentation: Created separate paths for SageMaker administrators and IAM Identity Center administrators, recognizing that setup requires collaboration between different teams with different expertise.
  • Architecture-first explanation: Started with clear architecture documentation to help readers understand how identity context flows between services before diving into configuration steps.
  • Service-specific integration guides: Created dedicated pages for each connected service (S3 Access Grants, EMR, EMR Serverless, Redshift Data API, Lake Formation, Athena) with service-specific prerequisites and configuration.
  • Security and compliance focus: Emphasized auditing capabilities with CloudTrail integration, helping enterprises meet compliance requirements.
  • Edge case documentation: Documented user background sessions and other advanced scenarios that enterprise users would encounter.

Documentation Structure

I designed an information architecture that guides administrators through the complex setup process:

  • Trusted identity propagation — Overview and use cases
    • Architecture and compatibility — Technical architecture, prerequisites
    • Set up — Configuration steps for both SageMaker and IAM Identity Center
    • Audit with CloudTrail — Compliance and monitoring
    • User background sessions — Advanced session management
    • Connect with other AWS services — Integration guides
      • Connect to Amazon S3 Access Grants
        • Studio JupyterLab notebooks with Amazon S3 Access Grants
        • Training and Processing jobs with Amazon S3 Access Grants
      • Connect to Amazon EMR
      • Connect to EMR Serverless
      • Connect to Redshift Data API
      • Connect to Lake Formation and Athena

Technical Depth

This documentation required deep understanding of enterprise identity management and cross-service AWS architecture:

  • IAM Identity Center integration: Documented how user attributes and group associations flow from Identity Center to connected services
  • Token exchange mechanisms: Explained how identity context is propagated without requiring users to re-authenticate
  • Service-specific authorization: Each connected service has different authorization models that needed to be documented clearly
  • CloudTrail event structure: Documented how to interpret audit logs to track user-specific actions across services

Connected Services Documentation

A major component of this work was documenting integrations with multiple AWS services, each with unique requirements:

Service Integrations Documented

  • Amazon S3 Access Grants: Fine-grained data access for notebooks, training jobs, and processing jobs
  • Amazon EMR: Cluster-based big data processing with identity propagation
  • EMR Serverless: Serverless Spark and Hive with user identity context
  • Redshift Data API: Data warehouse queries with user-level auditing
  • Lake Formation and Athena: Data lake access with fine-grained permissions

Enterprise Focus

This documentation was specifically designed for enterprise customers with complex identity and compliance requirements:

  • Compliance enablement: Detailed CloudTrail integration for audit trails
  • Least privilege support: User attribute-based access control
  • Cross-team workflows: Clear handoff points between admin teams
  • Scalability considerations: Documentation for large-scale deployments

What This Demonstrates

Ability to document complex enterprise security features spanning multiple AWS services. Experience with identity management, IAM, and cross-service authorization patterns. Skill in creating documentation for multiple technical audiences (SageMaker admins, IAM Identity Center admins, security engineers). Understanding of enterprise compliance and auditing requirements.