Warming up to Iceberg: 2026 makes Iceberg easier
What if you can run dbt without a Warehouse? In this talk we'll cover how you can load data to Snowflake managed (Horizon) Iceberg tables, transform the data using dbt-duckdb, all without turning on a Snowflake Virtual Warehouse. If you've been hesitant about Iceberg because of the setup complexity, we're going to cover recent improvements to DuckDB, dbt Fusion, and Snowflake's catalog that make this process easier than you think. And the best part: the moment your data lands, it's fully governed by Snowflake. RBAC, masking, lineage, all included.
Stub Snowflake for Databricks, etc. The examples in this talk focus on Snowflake but the ideas are cross-platform.
We'll also touch on the most up and coming Iceberg alternative, DuckLake, and when to reach for each tool.
Check out more sessions
- Breakout session
Meta:Context: a business context schema in dbt's semantic layer
Keith Binkly / Green Dot CorporationView session - Breakout session
Beyond the dashboard: an MCP-connected AI agent that reads your dbt lineage
Mars DauerView session - Breakout session
Optimizing your runs for lower compute, fresher data, and faster iteration with dbt State
Reuben McCreanor / dbt LabsIaroslav Zeigerman / FivetranView session
