Overview
This major insurance company, with more than $90 billion in annual premiums, has numerous use cases for fast, reliable data processing and analytics, including:
- Compliance with IFRS 17, one of the most significant challenges to insurance accounting in more than 20 years.
- Processing billions of insurance policies to assign the costs incurred for each policy across different areas.
Existing architecture failing the business
The company’s patchwork of legacy data warehouses were too slow, inflexible, unreliable, and expensive to support these use cases, slowing the pace of business and limiting the amount of accessible data:
- Policy expenses could only be subdivided four ways, and multiple days were required to process the entire portfolio.
- Insurance underwriting was too slow due to latency when running ad hoc queries, with only a subset of data available for analysis.
- The solution lacked the performance and scalability needed to meet IFRS 17 compliance requirements.
- Claims ratio calculation took two full days to complete for the full portfolio.
Modernizing for price/performance at scale with Yellowbrick
After extensive testing and evaluation, the company selected Yellowbrick over Greenplum and others for its data warehouse transformation project — with queries running 400X faster on Yellowbrick. During testing, Yellowbrick proved its ability to reliably process and analyze billions of rows of data (including numerous complex table associations), providing price/performance at scale, data consistency, and 24/7 availability.
Today, the company is using Yellowbrick to bring instant insights to asset management, annuity insurance underwriting, life insurance marketing campaigns, and HR (for 2 million employees).
Results include:
- Hundreds of underwriters can now create policies more quickly and accurately for customers, contributing to a reduction in average turnaround time from 3.8 days to 10 minutes
- Capabilities now in place for compliance with IFRS 17, with the company being the first in its country to meet that milestone
- Policy expense processing now completes in 2 hours (versus several days), with costs subdivided into hundreds of different areas to drastically improve management accounting decisions
- Claim ratio processing completes in 2 hours (versus 2 days), while concurrently running other processes without performance impact