Recommendation System for Digital Gaming
Objective
A leading ASEAN electronic gaming company needed an advanced Data Lake solution to consolidate user data from various SQL/NoSQL databases and large unstructured data streams. They also required a game recommendation engine based on real-time game data for increased game engagement. The key challenge was the size, type (unstructured), and volume of data.
About the Business
Smart Data Integration for Enhanced Engagement
The gaming company partnered with DataLens to implement a scalable data infrastructure, ensuring seamless data integration and analytics capabilities to enhance game engagement.
Solution Highlights
-
Change Data Capture (CDC)
Automated real-time data capture from MongoDB and ScyllaDB.
-
Web Scraping
Extracting live data from online game platforms.
-
Data Quality Checks
Delta Live tables for validation and error handling.
-
Data Lake Storage:
Accommodating diverse and voluminous data.
-
Security and Compliance
Data encryption and RBAC-based access governance.
-
Cloud-based ML Services
Developing and deploying the recommendation engine.
Key Outcomes
-
Data Pipeline Efficiency
Reduction in the time required for data ingestion and processing tasks by 40% and decrease the time-to-insight for analytics by 50% within the first 6 months
-
Recommendation Engine Performance
Increase click-through rate (CTR) on game recommendations by 30% and conversion rate by 15%
-
Analytics-Driven Decision Making
Utilize insights derived from the data lake to inform over 25 significant product or operational decisions within the first 9 months