GitLake: Git-for-data for the agentic lakehouse
For data engineers and AI agents managing lakehouse workflows, GitLake provides atomic, branch-based operations to ensure consistency and auditability.
GitLake introduces a Git-like version control system for lakehouses, enabling agents to work on isolated branches with atomic commits and merges. The system is validated through production deployment and a formal Alloy model.
We present GitLake, a Git-for-data design for an agent-first lakehouse. The system lifts single-table Iceberg snapshots into lakehouse-wide commits, branches, and merges, letting agents work on isolated branches while humans review and publish changes. Pipelines run on temporary branches and publish through a final merge, so all outputs become visible atomically or none do. Finally, we report production lessons as well as correctness insights from a preliminary Alloy model of our core abstractions.