Rocky

A compiled SQL transformation engine in Rust. Type-safe compilation, column-level lineage, and an AI intent layer — for data pipelines that don't break.

Get Started in 60 Seconds View on GitHub

Type-Safe Compiler

Column-level type inference and data contracts catch errors before execution.

Pure SQL or DSL

Write standard SQL with TOML config, or use the Rocky DSL for NULL-safe operators, windows, and pattern matching.

AI Intent Layer

Store intent as metadata. AI explains models, syncs them when upstream schemas change, and generates tests.

VS Code Extension

Full language server: completion, hover, go-to-definition, rename, code actions, and lineage visualization.

Quality Checks

Inline pipeline checks (row count, freshness, null rate, anomaly) plus model-level assertions with severity and quarantine.

Dagster Integration

Auto-discover assets, emit materializations, and translate check results — all via dagster-rocky.

Get Started in 60 Seconds

curl -fsSL https://raw.githubusercontent.com/rocky-data/rocky/main/engine/install.sh | bash
rocky playground my-first-project
cd my-first-project
rocky compile    # type-check
rocky test       # run assertions locally
rocky run        # execute the pipeline

No credentials needed — the playground is DuckDB-backed.

Coming from dbt? Import your dbt project, convert tests to contracts, and adopt Rocky incrementally.

Using Dagster? Auto-discovery, rich metadata, and check-result integration via dagster-rocky.