Governed by default: how data teams at Nordstrom turn dbt governance into an AI advantage
One question reframed how Nordstrom thinks about dbt: "What would it take for our data to answer questions on its own, and hold up under interrogation?"
That shift turned dbt from pipeline plumbing into the control surface for reliable AI. This session covers how Nordstrom's retail data teams are using dbt governance and semantics to build an AI-ready analytics foundation at scale: treating freshness as an AI SLA, using tests, contracts, and lineage as guardrails, and designing a "Data Concierge" conversational agent that answers merchants' questions in plain English against governed data.
You'll get a realistic view of the challenges, tradeoffs, and patterns a large enterprise is using to make its analytics platform ready for LLMs, not as slides, but as production decisions made by a team doing this work right now.
Check out more sessions
- Roundtable
Product roundtable: dbt Wizard: your AI teammate for data development
Ryan Eakman / FivetranView session - Breakout session
From 14-hour batches and poor documentation to AI-ready data: Mitti's dbt rebuild
Yuna (Yunnan) Tang / MittiThiago Baldim / MittiView session - Breakout session
All aboard: Rebuilding dbt in the open
Hope Watson / dbt LabsView session
