Lamin

Lamin is an open data platform for R&D teams that need to trace, query and govern data used by machine learning models and AI agents. Its core is LaminDB, an open source database and lakehouse that records lineage across agent sessions, notebooks, scripts and workflows, so results can be traced back to the data, code and compute environment that produced them. The company was founded in 2022 and has offices in Munich and New York City. It is aimed at computational biology, biotech and pharma teams as well as data scientists in other fields. The platform handles datasets in any storage format, including parquet, zarr and AnnData, with time travel, schema evolution and ACID guarantees, and works with engines such as Polars and DuckDB. Further features include validation and annotation of FAIR datasets, support for ontologies, a lab information management and electronic lab notebook layer, and branching and versioning of data and code for governance. It installs from PyPI or CRAN without Docker or a separate backend, and connects to local, S3 and GCP storage with Postgres or SQLite. The Free plan costs $0 per month. Pro is $20 per month billed annually and adds one hosted database, 10 GB of hosted storage and 10 GB of egress. Team is $200 per seat per month billed annually and adds an organizational account, ten hosted databases, single sign on, audit logs, fine grained permissions and a SOC2 report. Enterprise is custom priced and adds full on premise deployment in the customer's AWS account and consulting services. Academic institutions and public data initiatives are offered free and reduced pricing. The listing sits in the coding category, though the product is a data platform.