Polnor.
CRO

Run the study across every sponsor, without ever pooling their data.

You operate multi-sponsor, multi-site studies where no one will hand over their raw patient data, and rightly so. Polnor runs the same protocol inside each sponsor's own cloud, on an HDS-certified host, and lets you compare aggregate results across sites. You are the operator, not the custodian.

A worked case

One protocol, three sponsors, no data ever pooled.

A multi-site study run for three sponsors who will not share raw data with each other or with you. You define the cohort once, the same standardized pipeline runs in each sponsor's own environment, and only aggregate counts come back to be compared. Not a single patient row leaves any sponsor's cloud.

app.polnor.net / cro-multisite● health-blind
cro_multisite › omop › eligibility_cohort.sql
▶ Run per sponsor-- same protocol, runs in each sponsor's cloud SELECT count(DISTINCT person_id) n FROM condition_occurrence c JOIN drug_exposure d USING (person_id) WHERE condition_concept = 'endpoint' AND d.ingredient = 'X';
aggregate across sponsors · 3 sites0 rows exported ✓
siteeligible nevent ratecontributes
Sponsor A12,4803.0%counts only
Sponsor B9,2152.7%counts only
Sponsor C15,9023.4%counts only
pooled estimate37,5973.1%ref

Illustrative interface and figures, shown to explain the workflow. Not results from a Polnor customer.

The features behind it

Standardized in every environment, so results are comparable.

The same automated pipeline runs identically in each sponsor's cloud: ingest, de-identify, standardize, quality-check. Because every site is prepared the same way, the aggregate numbers you compare actually mean the same thing.

Automated · ingestion

Each sponsor's data in, at the FHIR standard

Inside each sponsor's environment, claims, EHR extracts and lab feeds are pulled and normalized to FHIR R4 automatically, one typed table per resource. No bespoke ETL per sponsor, and no data crossing an environment boundary.

Connectors + automated $export pull, per sponsor
Typed bronze tables, refreshed on schedule
ingestion · per sponsorauto
sponsorresourcesstatus
Sponsor AFHIR R4✓ auto
Sponsor BFHIR R4✓ auto
Sponsor CFHIR R4✓ auto
Automated · de-identification & OMOP

Identical de-identification and OMOP mapping at every site

Direct identifiers are masked and dates shifted by the same preset in each environment, then the data is converted to the OMOP CDM with OHDSI concept mapping and clinical quality control. Standardizing every site the same way is what makes the cross-sponsor comparison valid, not just convenient.

Same GDPR presets applied automatically in every environment
Automated OMOP conversion, concept mapping and QC
omop · mappingauto
source→ OMOP conceptstd
atc:Xingredient XS
icd10:I26Pulmonary embolismS
loinc:718-7HemoglobinS
You · analysis & models

Define the cohort once, compare the aggregates

This is the part your study team owns: write the eligibility and endpoint logic once, dispatch it to every sponsor's environment, and read back only aggregate counts to compare. If a model is in scope, train and track it with MLflow next to the data. Nothing patient-level is ever exported.

Cohort builder, SQL and notebooks on each sponsor's OMOP tables
MLflow-compatible tracking and serving in each sponsor's cloud
aggregate · across sponsorscounts
siteeligible nexported
Sponsor A12,4800 rows
Sponsor B9,2150 rows
Sponsor C15,9020 rows
Almost everything is automated

The platform does the plumbing. Your team runs the study.

The line is deliberate: the repetitive, error-prone, compliance-heavy work runs itself in every environment, so your study team spends its time on protocol and interpretation, not on reconciling three incompatible datasets.

Automated by Polnor
FHIR ingestion · refresh, per sponsor
GDPR de-identification
OMOP conversion · concept mapping
clinical quality control
feature materialization · lineage · audit log
Your team decides
the protocol & endpoints
eligibility & cohort definition
which aggregates to compare
the analysis, and the interpretation
Operator, not custodian

Nothing to exfiltrate, because you never hold the data.

Each sponsor's environment is isolated: their data never mixes with another sponsor's and is never pooled. Every access to patient data is logged, PHI is classified, and export follows EHDS-aligned manifests. Because the control plane is health-blind and the data stays in each sponsor's HDS-certified cloud, there is nothing on your side to leak.

See it on a study like yours

One protocol, every sponsor, no data pooled.

Book a 30-minute demo. We run the automated pipeline across a mock multi-sponsor study and show you exactly how the same standardized logic executes in each environment while only aggregate results come back.