We modernized Disney’s Operation Metadata Store (OMS) by transitioning from a SaaS-based metadata platform to a fully AWS-native architecture with a custom implementation, aligning the initiative under the Re-purchase category of AWS modernization.

The modernization replaced externally managed SaaS dependencies with a purpose-built, cloud-native solution using AWS managed and serverless services. This shift improved architectural control, cost efficiency, scalability, and operational resilience, while enabling tighter integration with Disney’s data and analytics ecosystem.

The Walt Disney Company operates large-scale data platforms supporting analytics, governance, and metadata-driven insights across its digital ecosystem.

OMS serves as a centralized metadata intelligence layer, providing visibility and analytics across datasets and tables used by engineering and analytics teams. The platform underpins metadata discovery, lineage analysis, and interactive analytics, making it a critical foundation for data-driven operations.

The legacy OMS capability was delivered through a SaaS-based product, which introduced several strategic and operational constraints:

• Limited flexibility in extending or customizing metadata workflows
• Dependency on third-party release cycles and feature roadmaps
• Higher total cost of ownership at scale
• Challenges integrating deeply with AWS-native analytics and governance services
• Reduced architectural control over resiliency and security posture

As metadata usage and analytics needs expanded, Disney required a platform that could scale seamlessly, integrate natively with AWS data services, and provide long-term cost and architectural efficiency.

We adopted a Re-purchase modernization approach, replacing the SaaS product with an AWS-native implementation, purpose-built to meet Disney’s functional, scalability, and governance requirements.

The strategy focused on:

• Leveraging AWS managed and serverless services wherever possible
• Designing for inherent scalability, resiliency, and fault tolerance
• Reducing operational overhead by minimizing infrastructure management
• Optimizing cost by selecting service models aligned with usage patterns
• Enabling future extensibility without vendor lock-in

We implemented OMS using a combination of AWS serverless and managed services, delivering a modular and highly scalable metadata platform:

• Metadata Analytics & Search delivered through Amazon OpenSearch and OpenSearch Dashboards
• Lineage and relationship analysis implemented using AWS Neptune
• Interactive analytics enabled via AWS Glue Data Catalog and Amazon Athena
• API and application layer built using AWS AppSync, Lambda, and Amazon ECS on Fargate
• Frontend and content delivery served via Amazon CloudFront and Amazon S3
• Authentication and access control implemented using Amazon Cognito

This architecture provides a unified metadata experience while maintaining loose coupling between analytics, lineage, and discovery components.

OMS was designed using AWS-managed and serverless services, which inherently provide high availability and fault tolerance:

• Services such as CloudFront, S3, AppSync, Lambda, Athena, Glue Data Catalog, ECS Fargate, Step Functions, and Neptune Serverless automatically handle scaling and failover
• Multi-AZ resiliency is built into the platform without requiring custom failover mechanisms
• No additional infrastructure is required to manage availability or scaling for the majority of the stack

Amazon OpenSearch is the only component that requires explicit capacity management. Given the platform’s usage profile-primarily serving a few hundred data engineers—we intentionally selected provisioned OpenSearch over OpenSearch Serverless to optimize cost.

• A small, stable cluster footprint meets current workload demands
• Capacity is manually managed due to predictable usage patterns
• This approach balances operational simplicity with cost efficiency

• The solution is deployed within a single AWS region
• Multi-AZ deployments across serverless services provide sufficient resiliency
• Multi-region deployment was intentionally excluded, as it was not required for the OMS workload profile

This design ensures reliability while avoiding unnecessary architectural complexity.

The modernization delivered clear strategic and operational benefits:

• Eliminated reliance on third-party SaaS metadata platforms
• Improved architectural control and extensibility
• Reduced operational overhead through AWS-managed services
• Optimized costs by aligning service choices with usage patterns
• Enabled deeper integration with AWS-native analytics and governance services
• Improved platform resilience through built-in AWS high availability

By adopting a Re-purchase modernization model - replacing SaaS with AWS-native services - we achieved:

• Greater long-term cost efficiency
• Improved security and compliance alignment
• Faster innovation through native AWS integrations
• Reduced vendor dependency and increased architectural ownership

This approach delivered a future-ready metadata platform aligned with Disney’s broader cloud and data strategy.

We successfully modernized Disney’s Operation Metadata Store by re-purchasing a SaaS capability with a fully AWS-native, custom-built solution. The transformation improved scalability, resiliency, cost efficiency, and architectural control, while establishing a strong foundation for future metadata analytics and governance capabilities.

This engagement demonstrates AWS best practices in application modernization, re-purchase strategy execution, serverless architecture, and operational excellence.