Lamin Labs

Lamin is an open data platform for biology, providing a lineage-native lakehouse that serves as a programmable context and memory layer for biological R&D. Founded in 2022 and headquartered in Munich and New York City, Lamin was built to solve a critical bottleneck in AI-driven biological research: closing the feedback loop across heterogeneous data modalities at scale. The platform enables teams to query, trace, and validate complex biological datasets, building organizational memory that compounds over time and supports the iterative cycle between laboratory results and machine learning models.

Lamin serves thousands of scientists across academia, biotech, and global pharma. Its core product, LaminDB, evolved from a "git for R&D data" concept into a full-featured, open-source, zero-lock-in data framework that scales from personal projects to pharma-scale enterprise deployments — much like git itself.

Core Platform Capabilities

• Data Lineage: Track where data came from and what it is used for with a single line of code. Lineage tracking supports notebooks, scripts, pipelines, and functions, and works across Python, R, shell, Nextflow, and SQL environments.
• Lakehouse with Bio-Format Support: Query and batch-load datasets at scale using lakehouse architecture. Supports a wide range of table and array formats including .parquet, .zarr, AnnData, and SpatialData. Dataset features and schemas are managed as metadata in Postgres or SQLite.
• Registries and Sheets (LIMS): Manage metadata in relational sheets that stay in sync with datasets in storage. Provides a unified Python/R class interface with built-in ontologies, project management, and change management. Supports experiments, samples, notes, reports, and more.
• Data Integrity: Use schemas to enforce consistency across data assets. Annotate datasets with a single line of code, supporting formats such as sheets, .parquet, .csv, .zarr, AnnData, and SpatialData.
• Organizational Memory: As teams and agents work, data, models, and reports are mapped into the lakehouse, building recursively queryable memory and training data that grows and compounds over time.

Deployment and Infrastructure

• Zero Lock-In Architecture: Lamin is designed so that organizations retain full admin control over their own infrastructure. Fine-grained permissions for both humans and agents can be managed with SaaS-like simplicity directly at the database and storage level.
• Database Support: Compatible with both Postgres and SQLite backends.
• Storage Integrations: Supports a broad range of storage options including local file systems, AWS S3, Google Cloud Platform (GCP), Microsoft Azure, and Cloudflare R2, among others.
• Open Source: LaminDB is available as an open-source package installable via pip install lamindb, with support for Python and R.

Team and Background

• Alex Wolf (Co-Founder & CEO) created Scanpy and led the build-up of Cellarity's compute platform, with over two decades of R&D software experience and more than 20,000 citations on Google Scholar.
• Sunny Sun (Co-Founder & President) previously served as Head of Computational Biology at Cellarity, with a background in genome engineering and cell biology.
• Frederic Enard (Co-CTO & Founding Engineer) brings backend and infrastructure expertise, previously serving as CTO at AI startup Kicck and with 4.5 years in enterprise data engineering at TF1.
• Sergei Rybakov (Founding Principal Engineer) focuses on the data layer and open core, with early contributions to the scanpy and anndata teams and a background in machine learning and computational biology.
• Richard Sriworarat (Founding Full Stack Engineer) handles frontend and science, created Samui, and holds a Neuroscience PhD from Johns Hopkins.

Investors and Backers

• Lamin is backed by notable investors including Aaron Kimball (Co-Founder, WibiData), Alec Nielsen (Co-Founder & CEO, Asimov), Jeff Hammerbacher (Co-Founder, Cloudera), and Oskari Saarenmaa (Co-Founder & CEO, Aiven).

Lamin provides a reliable, open, and scalable foundation for data-driven biological research, enabling scientific teams to build long-term institutional memory while maintaining full control over their data infrastructure.