Business Situation
- A US-headquartered insurance company faced significant challenges with its crediting rate process.
- The manual, time-consuming, and error-prone process involved multiple teams and extensive calculations in spreadsheets.
- It took 7–8 days to complete each cycle, causing delays in rate updates.
- Data scattered across email threads and individual spreadsheets led to inconsistency and loss.
- A lack of real-time visibility into the process status made tracking changes and maintaining an audit trail difficult.
SGA Approach
Technology
- Leveraged AWS services, including Redshift, Glue, Lambda, Step Functions, and simple email service (SES)
- Implemented data versioning and partitioning strategies for optimal query performance
AI
- Utilized process mining techniques to map existing process flow
- Implemented a custom business rules engine (BRE) for complex decision-making automation
Data
- Created a centralized data repository with version control and audit trails
- Implemented a three-tier data architecture (bronze, silver, gold) for efficient data management
Key Takeaways
- Implementing a centralized data repository using AWS Redshift with extract-transform-load (ETL) pipelines and a three-tier data architecture resulted in a 95% improvement in turnaround time. From 7 to 8 days, it decreased to 2 to 3 hours.
- Deploying a custom BRE using AWS Lambda functions facilitated automated, efficient decision-making.
- Developing interactive dashboards using AWS Step Functions enabled real-time monitoring and process orchestration.
- Automating 80+ hours of manual effort per cycle improved cost optimization and achieved 100% data consistency and availability.
