Cohorts
Built inside the hospital. Never export raw data. The substrate carries the cohort, not a CSV.
Research
Research, with the patient still in the loop.
De-identified cohorts. Reproducible trials. Sovereign data governance. The hospital builds the cohort, runs the protocol, publishes the result, and never exports raw data.
How research changes
Trials
Protocol fidelity, audited end to end. Every step on the chain.
Publication
Reproducible artifacts attached to every result. Citation discipline. No ghost data.
Governance
Sovereign jurisdiction. Ethics review on file. Patient consent surfaced in the chain.
The reproducibility contract, in research
Same cohort definition, same protocol, same result. The artifact survives the trial: cohort SQL, model card, protocol PDF, signed consent ledger, audit excerpt. A reviewer with the same hospital deployment reproduces the result, not the screenshot.
- 0
- raw exports
- 100%
- protocol on chain
- 1
- sovereign jurisdiction
- 1
- reproducibility contract
Why the hospital is the lab
The research bench is the unit. The senior consultant who runs the trial signs the protocol, the inclusion, the analysis, the publication. The instrument is the substrate, not the lab notebook.