dbt Labs
at Snowflake
Summit

In this session, Christine Kwon, Senior Director of Data at Fanatics, and Brett Pendleton, Manager of Analytics Engineering at Fanatics, share how they used dbt to define and govern Fanatics’ metric source of truth using scalable, metadata-driven models (YAML-based definitions, testing and documentation). They will walk through how those curated models power Snowflake semantic views and Snowflake Intelligence, enabling business users to ask trusted, governed questions in natural language.

In 90 minutes, you'll go hands-on building a production-ready AI chatbot — taking raw data from ingestion all the way to an experience powered by Snowflake Intelligence using Fivetran and dbt Platform. Guided step-by-step, this lab is designed for all skill levels. Whether you're new to the modern data stack or a seasoned practitioner, you'll walk away with working knowledge of Snowflake Cortex AI and a repeatable blueprint you can take back to your organization.

CarGurus runs a hub-and-spoke data organization — a central team with embedded analytics teams across the business. That model created friction: PR bottlenecks, inconsistent standards and inefficient backfills consuming engineering capacity. This session covers how CarGurus deployed the dbt Fusion engine on Snowflake to intelligently optimize builds, reclaim 24,000 annual engineering hours and maximize business value, achieving ROI in just five months. We also cover how that recovered capacity is fueling AI experimentation with Snowflake Cortex AI, Streamlit and the dbt semantic layer.

Agents can generate outputs in seconds, but without visibility into their reasoning, data sources and logic, teams can't trust them in production. This session shows how dbt powers trusted Cortex Agents through a governed context layer and high-precision retrieval — ensuring agents start from the right models and metrics, cite their sources, align to approved definitions and fail safely when context is unavailable. Using dbt agent workflows, you can deploy autonomous agents with traceable, reviewable outputs. Leave with a practical framework for moving from agent demos to production-ready deployments you can trust.

The dbt Fusion engine is here, and if you're running dbt on Snowflake, this is the upgrade you've been waiting for. With native Snowflake ADBC/Arrow integration, Fusion delivers faster query previews, snappier CI runs, and IDE-quality autocomplete and inline errors, all while generating rich semantic metadata that makes your dbt project understandable to both humans and LLMs. In this session, we'll walk through enabling Fusion in dbt on Snowflake and show you exactly what changes in your development experience — including the gotchas — to know before you flip the switch in production.

In 90 minutes, you'll go hands-on building a production-ready AI chatbot — taking raw data from ingestion all the way to an experience powered by Snowflake Intelligence using Fivetran and dbt Platform. Guided step by step, this lab is designed for all skill levels. Whether you're new to the modern data stack or a seasoned practitioner, you'll walk away with working knowledge of Snowflake Cortex AI and a repeatable blueprint you can take back to your organization.