From dashboards to data products: how dbt transforms analytics teams

Many analytics teams start the same way: as reporting factories. SQL logic gets duplicated, metrics drift apart, and trust in data erodes. As AI enters the analytics stack, the stakes are even higher: AI is only as reliable as the data it's built on. 

Over seven years, I've worked with dbt across three organizations at different stages of maturity: discovering dbt as an analyst and transitioning into analytics engineering; scaling it through multi-repo and data mesh architectures; and now leading a full team transformation toward data as a product at CtM. 

I'll share practical lessons from all three stages: building a testing culture, scaling ownership across domains, and shifting from dashboards as outputs to reliable data models as products. Because that foundation is what AI and agentic workflows actually need to work.