Build AI-ready data foundations on ClickHouse with the dbt Fusion engine

The dbt Fusion engine is launching in Private Preview for ClickHouse, and ClickHouse will be among the first partner-developed Fusion adapters. Initially available via the dbt VS Code extension and dbt platform, so you can bring real-time analytics and AI workloads under on trusted control plane.

Fusion brings dbt’s transformation standard to ClickHouse’s near-real-time read/write platform, with governance, lineage, and column-level metadata automatically applied to every transformation. This is exactly what modern analytics and AI workloads need.

This is a real preview with hands-on support from the dbt Labs and ClickHouse teams. Sign up to request early access to Private Preview: a direct feedback channel to product and engineering and a seat at the table as we build toward GA.

Orders of magnitude faster development

With parse times up to 30x faster than dbt Core, Fusion executes large dbt projects in milliseconds instead of minutes. On ClickHouse, that speed compounds: real-time SQL validation locally, state-aware orchestration that skips what isn't new, and ~30% average compute savings without changing how you write SQL.

Governed transformation, built in

Define you business logic once in dbt; govern it centrally across every ClickHouse workload. Column-level lineage, impact analysis, and sensitivity tagging are automatically applied to every transformation — so business definitions stop diverging across dashboards, microservices, and ad-hoc SQL.

A data foundation built for AI

Modern analytics and AI projects demand that teams can transform and serve data seamlessly across their stack. Fusion on ClickHouse means your transformation layer is production-grade, context-rich, and trustworthy — paired with dbt's MCP server, it gives AI agents the governed entry point they need to do trusted work on your ClickHouse data.

What you can expect from private preview:

  • Early access to Fusion on ClickHouse before GA
  • Hands-on onboarding and support from the dbt Labs team
  • Direct feedback loop with product and engineering
  • First look at new capabilities as they ship