From mainframe to AI-powered maturity: Verisk's data modernization journey

Verisk Analytics’ underwriting data team described a four-stage modernization of an insurance data platform that processes billions of records from hundreds of insurers and must maintain high accuracy for actuarial and regulatory use. Its longstanding mainframe reliably executed decades of business rules but obscured data lineage and discouraged changes, while an initial migration to Snowflake and Jupyter notebooks improved visibility and version control but was unsuitable for daily automation. Native Snowflake tools enabled more frequent refreshes, yet created a difficult-to-manage network of tasks, streams, tables, and manual deployment processes. Adopting dbt centralized version-controlled transformations, peer reviews, automated testing, documentation, lineage, and environment management, while naming conventions and custom macros addressed Snowflake schema and model-name constraints. This governed foundation also supports human-reviewed AI applications, including automated documentation propagation, context-aware SQL generation, pull-request standards checks, and faster debugging through log analysis. The team noted that organizational adoption requires staff to develop more technical skills and that efficiently updating more than a decade of historical data remains an unresolved scaling challenge.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.