Stop feeding your agent the whole repo: token-efficient code intelligence for dbt projects
Breakout sessionBuilding agentic data workflowsIntermediateData leadersData practitionersAll industries
Ramp's data team has been deep in the AI-first transition: building internal agents, rethinking BI, and stepping into knowledge management. This session shares what's actually working.
Topics include: evolving our internal data analyst agent Ramp Research into a coding agent running in a sandbox environment; Wavedash, our new AI-first internal BI tool built around the requirements generic tools couldn't meet; Ramp Wiki, how the data team stepped into prose-based knowledge management; and what we've learned tracking AI usage across Ramp's product and internal tooling.
Expect honest observations, not a polished case study. The space is moving too fast for that.
Check out more sessions
- Breakout session
One engine, new era: Inside dbt Core v2.0
Hope Watson / dbt LabsView session - Breakout session
Governed by default: how data teams at Nordstrom turn dbt governance into an AI advantage
Priya Tanwar / NordstromNadine Bruxel / NordstromView session - Hands-on lab
Scaling trusted self-service for dbt stakeholders
Kyle Tuft / dbt LabsMatteo Dijoux / dbt LabsView session
