# Rocky

> **TL;DR.** Rocky is a Rust-based SQL transformation engine and control plane for data warehouse pipelines.

- **Category:** Data Engineering / Developer Tools / Infrastructure
- **Stage:** validating
- **Age:** 64 days
- **Origin date:** 2026-04-13
- **First detected:** 2026-04-29
- **Canonical URL:** https://earlyterms.com/term/rocky
- **Sources:** 4 primary URLs

## Definition

Rocky is a Rust-based SQL transformation engine and control plane for data warehouse pipelines. It layers compile-time safety, column-level lineage, named branches, and per-model cost attribution on top of existing warehouses like Databricks or Snowflake — without replacing storage or compute.

The [GitHub repo](https://github.com/rocky-data/rocky) went public on April 13, 2026, followed by `dagster-rocky` on PyPI the next day and a [Show HN post](https://news.ycombinator.com/item?id=47935246) on April 28 that drew commentary from an engineer at dbt Labs. Rocky positions as a type-safe replacement for dbt and SQLMesh with 68+ releases in its first two weeks.

## Example

A data team runs `rocky branch staging && rocky run` to materialize pipeline models in an isolated schema before promoting to production. Column-level lineage shows which downstream fact tables are affected by a single upstream type change — catching blast radius before a single production row is written.

## Analogy

Think of it as a compiler and Git branching system combined, but for SQL warehouse pipelines.

## Why it's emerging now

The SQLMesh acquisition left data engineers without a credible indie alternative to dbt, and dbt's Jinja-heavy approach has long frustrated compile-time-safety advocates. Rocky launched on April 13, 2026 with Rust-speed compilation, column-level lineage, and named branch isolation — and a dbt Labs engineer publicly acknowledged both features as missing from dbt within 24 hours of the Show HN.

## Related terms

- *competitor:* dbt
- *competitor:* SQLMesh
- *competitor:* dbt Fusion
- *related:* column-level lineage
- *related:* data contracts
- *parent:* warehouse pipelines
- *parent:* data pipeline
- *related:* Dagster
- *related:* schema drift
- *related:* ggsql

## Sources

1. [Rocky — GitHub repository (rocky-data/rocky)](https://github.com/rocky-data/rocky)
2. [Rocky — official documentation site](https://rocky-data.dev/)
3. [Show HN: Rocky — Rust SQL engine with branches, replay, column lineage](https://news.ycombinator.com/item?id=47935246)
4. [dagster-rocky on PyPI — version history and integration description](https://pypi.org/project/dagster-rocky/)

---
_Generated by EarlyTerms · https://earlyterms.com/term/rocky_