From first models to AI-powered maturity: scaling data engineering with dbt at Verisk Underwriting

How do you evolve from early dbt adoption to a mature, automated, AI-augmented data platform? 

dbt started at Verisk Underwriting as a lightweight modeling tool. It's now a full data engineering platform, with AI accelerating development, lineage driving faster impact analysis, and intelligent orchestration replacing schedule-based rebuilds.

We'll walk through how we scaled dbt adoption across teams, used lineage to improve trust and speed up root-cause analysis, chose between dbt jobs and Snowflake Tasks for orchestration, and applied dbt Copilot to model development, YAML generation, and documentation. 

We'll share the tradeoffs we made, the patterns that held up at scale, and how AI is shifting data engineers toward higher-value design work, letting teams move faster with greater confidence and less manual overhead.