This article describes how Ruparupa, an Indonesian ecommerce platform, built a real-time data lake using AWS services to gain faster business insights.

• Ruparupa faced challenges with manual data processing taking 1-1.5 hours for ingestion
• Initial solution used AWS DMS and Aurora MySQL for real-time CDC updates to QuickSight dashboards
• Upgraded to mutable data lake using S3, AWS Glue, Apache Hudi, and Athena for scalability
• AWS Glue ETL jobs with Hudi connector merge hourly incremental updates from raw to transformed layer
• Secrets Manager stores configuration parameters for generic transformation jobs across multiple tables
• Insights now available hourly instead of weekly—168 times faster delivery
• QuickSight dashboards increased from 20 to 180 users across business units
• Analytics team grew from 1 to 7 people handling new use cases
• Implementation enabled faster decision-making and increased sales revenue

Ruparupa successfully transformed from manual weekly reporting to hourly automated insights, enabling broader organizational access and faster business decisions.