lakeFS

lakeFS is an open-source data versioning platform built on top of object storage systems such as Amazon S3, Google Cloud Storage, and Azure Blob Storage. The company was founded by Einat Orr and Oz Katz. lakeFS is headquartered in Tel Aviv.

The core product — also called lakeFS — applies Git-like version control semantics (branches, commits, merges, rollbacks) to data lakes stored on object storage. This solves a concrete problem for data engineering and ML teams: the inability to safely experiment with, reproduce, or roll back large datasets without duplicating data or disrupting production pipelines. The primary buyers are data engineers, ML engineers, and data platform teams at organizations running large-scale data lakes. lakeFS is distributed as open-source software, with a commercial managed cloud offering (lakeFS Cloud) for teams that want a hosted, enterprise-supported version.

The open-source lakeFS project is the flagship product, available on GitHub and usable as a self-hosted layer on top of S3-compatible storage. lakeFS Cloud is the managed SaaS version aimed at enterprise data teams. The platform integrates with common data ecosystem tools including Apache Spark, Apache Hive, Presto, dbt, and Delta Lake, and works natively with AWS, GCP, and Azure object storage.

The primary framing for lakeFS is "data version control" or "data lakehouse versioning." Direct conceptual competitors include Delta Lake (from Databricks) and Apache Iceberg, both of which offer table-format-level versioning rather than a Git-style branching layer across all file types. lakeFS differentiates by being format-agnostic and by applying branch/merge semantics at the storage level rather than requiring a specific table format. The open-source model also competes with internal tooling that data teams build themselves.

lakeFS operates out of Tel Aviv. The founders, Einat Orr and Oz Katz, are Israeli. R&D is centered in Israel.