Implementing Travel and Hospitality Data Mesh

Data flows into AWS through batch processing, real-time data, SFTP, and Internet of Things (IoT) sensors.

Data sources are managed by the business domain. Producers use organization-level blueprints for security, governance, and open standards. Producers build the operational data store by using Amazon DynamoDB and DynamoDB Accelerator (DAX) for consistent single-digit-millisecond latency.

Consumer domains process data sets from multiple producer domains based on business needs. Build data marts in Amazon Redshift and analytics in AWS Quick Sight.

Manage metadata through multiple services. Use AWS Glue for data cataloging. Store data lineage in Amazon Neptune. Store data contracts in DynamoDB.

Consumers search and filter data sets in the central data catalog. Filter by name, contents, sensitivity, or custom labels.

Lake Formation provides centralized management of security, governance, and auditing with fine-grained permissions. It provides automatic schema discovery and format conversion.

Perform machine learning (ML) by using Lake Formation. Use Amazon SageMaker AI for standard AI/ML models. Use Amazon Personalize for actionable insights.

Data flows into AWS through batch, real-time, SFTP, and IoT sensors.

Stage all batch and real-time data in Amazon S3.

AWS Glue crawler creates table metadata in the data catalog.

Use open standards to build the data lake. Use a read-pattern schema for raw and curated data.

Use purpose-built databases like DynamoDB and serverless architecture for microservices and events in the operational data store.

Build reportable data sets in Amazon S3. Use Amazon Redshift and Athena for analytics. For ad hoc requirements, use Athena with standard SQL.

Use SageMaker AI for AI/ML models.

Use a multi-account strategy for resource and security isolation.