The query moves. The data doesn’t.

Tamarin lets researchers run approved computation against clinical data while participating healthcare organizations maintain control over their records.

How It Works

Data becomes research-ready

Participating organizations map approved source data into a standardized Tamarin representation so research logic can operate consistently across different data environments.

The researcher brings the analysis

Researchers can submit Python-based queries, analytical code, models, or protocol logic.

Computation happens against the source

Approved research executes against the participating organization’s data environment rather than requiring the underlying patient-level dataset to be delivered to the researcher.

Only the authorized research output moves forward

Researchers receive approved results. When a study involves recruitment, participating healthcare organizations can use locally identified patient populations within the appropriate study workflow.

Standardize once. Query many ways.

Tamarin uses a standardized entity–attribute–value (EAV) representation to create a common research layer across heterogeneous healthcare data.

That lets researchers write flexible Python analyses without having to reinvent the underlying data integration for every study.

Bring Your Own Code

Tamarin is not limited to a fixed library of predefined queries. Researchers can use Python to express the analysis their study requires, subject to participating Data Owner approval and applicable study controls.

Privacy-preserving by design

Tamarin is designed to let organizations participate in research without making permanent patient-level data transfers the default mechanism for collaboration.

Ready to put the network to work?

Run an analysis, evaluate a protocol, validate a model, or discuss a clinical study using source-connected healthcare data.