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

DevFeed: [From mainframe to AI-powered maturity: Verisk's data modernization journey](<https://devfeed.tech/articles/from-mainframe-to-ai-powered-maturity-verisk-s-data-modernization-journey-101844.md>)

Original publisher: [Read original article](<https://www.getdbt.com/blog/verisk-mainframe-dbt-data-modernization-ai>)

Author: Daniel Poppy

Published: 2026-10-26T14:04:00Z

Content type: news

Language: en

Sources: [dbt Labs](<https://devfeed.tech/sources/dbt-labs.md>)

Topics: [Legacy Modernization](<https://devfeed.tech/topics/legacy-modernization.md>), [Data pipelines](<https://devfeed.tech/topics/data-pipelines.md>), [Mainframe](<https://devfeed.tech/topics/mainframe.md>), [AI-Ready Data](<https://devfeed.tech/topics/ai-ready-data.md>), [etl](<https://devfeed.tech/topics/etl.md>), [Pull Request](<https://devfeed.tech/topics/pull-request.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-powered](<https://devfeed.tech/tags/ai-powered.md>), [data](<https://devfeed.tech/tags/data.md>), [dbt](<https://devfeed.tech/tags/dbt.md>), [decades](<https://devfeed.tech/tags/decades.md>), [insurance](<https://devfeed.tech/tags/insurance.md>), [journey](<https://devfeed.tech/tags/journey.md>), [learn](<https://devfeed.tech/tags/learn.md>), [mainframe](<https://devfeed.tech/tags/mainframe.md>), [maturity](<https://devfeed.tech/tags/maturity.md>), [modernization](<https://devfeed.tech/tags/modernization.md>)

## AI overview

Verisk moved underwriting data from a decades-old mainframe to Snowflake notebooks, then native Snowflake objects, and finally dbt as each step exposed new limits. Notebooks improved visibility and version control but did not suit daily production; Snowflake streams, tasks, and dynamic tables enabled daily refreshes but left hundreds of objects to coordinate manually. dbt unified the workflow with reviewed, version-controlled models, lineage, and tests that stop bad data before it reaches actuaries. The team organized models by business line and layer, used a naming convention to ensure uniqueness, and added a macro to route models to developer or production schemas. This governed foundation also gives AI tools richer context for propagating documentation, generating SQL, checking pull requests, and debugging; people remain in the loop. The excerpt says incremental models are still catching up to actuaries' request for fully updated historical data spanning more than ten years.

## Source excerpt

Verisk modernized decades of insurance data, from mainframe to dbt, to get AI-ready.