This article describes how Amazon Finance Automation built a data mesh architecture using AWS Lake Formation and AWS Glue to enable distributed data ownership while centralizing governance across 850+ datasets.

• Data mesh architecture enables domain teams to own and govern their datasets independently
• Central governance account uses AWS Lake Formation for secure data catalog and access management
• Global catalog aggregates metadata from producer AWS Glue Data Catalogs across accounts
• Self-service tools automate producer onboarding, consumer access requests, and approval workflows
• Fine-grained access controls applied through Lake Formation permissions and AWS RAM sharing
• Metadata sync process continuously updates central catalog with schema and partition changes
• Backup repair service provides self-healing mechanism for reliability and resiliency
• Supports 300+ curated data products accessible via Athena, Redshift Spectrum, EMR, QuickSight

The data mesh eliminates duplicate pipelines, reduces data movement costs, and enables faster innovation through self-serve data access with centralized governance and security controls.