A hospital biobank corridor at dusk, rows of sealed sample freezers under low light.

One record per patient, for research and AI.

For research directors and registry leads: each patient's records joined inside your hospital, ready for EHDS duties from 2029.

Book a demo

One patient, six systems, no answer.

A researcher asks which patients have a scan, a slide, a genetic result, and a consented sample, and today that question is a project.

Six system owners asked, six exports matched by hand
Studies and data requests that start from scratch each time
AI projects that ask for your data to leave

Which patients have a scan, a slide, a genetic result, and a consented sample?

Health records
MRN 0088213HL7 feed
Laboratory
LIS-5521-BManual lookup
Imaging
ACC 24-118830CSV export
Pathology
S24-0417 A1Email attachment
Genomics
SEQ-7F3KNo link
Biobank
BB-000912

Illustrative. Sample identifiers.

Every source, one record per patient.

The Scientific and Research Data Hub joins the research records your hospital holds, system by system, into one record per patient, inside your walls: health records, laboratory results, imaging, digital pathology, genomics, biobank samples, trial data, physician observations, and follow-up outcomes. One permission system and one audit trail cover all of it.

Health records
Laboratory results
Imaging
Digital pathology
Genomics
Biobank samples
Trial data
Physician observations
Follow-up outcomes

One record

See the whole patient in one place.

Scans, slides, genetic results, and samples sit in the record they belong to, viewed and annotated under the same permissions.

Radiology in DICOM, viewed and annotated in the record
Whole slide pathology beside the diagnosis and molecular result
Biobank samples tied to consent and chain of custody
Laboratory results traced to the method and calibration behind them

A chest CT slice in the record's radiology viewer, one region outlined. — CT image: Anti-PD-1 Immunotherapy Lung, The Cancer Imaging Archive, CC BY 3.0, cropped.Slide image: CPTAC-LSCC, The Cancer Imaging Archive, CC BY 4.0, cropped.

Illustrative. Sample data.

AI that runs inside your hospital.

AI works where the records sit, on your own servers, and sees only what the person asking is allowed to see.

Your local AI model

It runs on your own servers, so questions, records, and answers stay inside the building.

Your researchers' models

Models your teams build or adopt can be deployed inside your hospital, to train and test where the records sit.

A hospital server room at night, racks receding into the dark, one warm light far down the aisle.

What your teams can ask.

Ask in plain language
Find patients across every record type with one plain question, answered only from records the person asking may see.
Size a study first
See how many patients match a study's criteria before anyone asks for an export.
Spot gaps earlyIn development
Find missing consent, unlinked samples, and uncoded records before a data request arrives.
Summarise a historyIn development
Pull one patient's history across departments together for a study review.
Draft dataset descriptionsIn development
Draft the dataset descriptions the EHDS asks for, from what you actually hold.

Connect your own tools through the open REST API, on your own infrastructure.

Nothing leaves without your approval.

Your records stay where your hospital puts them, in the form they were created, until it decides otherwise.

Request
Approval
Anonymised extract
Time-limited link
Release logged

Released only on approval

A request is approved by your hospital, the extract is anonymised, it travels by a time-limited link, and the release is logged.

Kept as captured

Records are kept as entered, never pre-summarised, so researchers combine them at analysis, and studies nobody has designed yet stay possible.

On-premise, hybrid, or cloud: your hospital chooses where the data hub runs.

Ready before the requests arrive.

Four demands are reaching European hospitals from 2029, and each needs records that are joined up and governed.

EHDS data duties
From 2029, describe datasets and answer requests within three months
Descriptions from what you hold, requests from one place
Digital maturity scores
Scores such as HIMSS EMRAM rate how well systems share data
One record across departments, the integration scores look for
AI on health data
The EHDS counts training and testing AI as research
Records already linked and governed for AI work
Research networks
Networks analyse data where it sits, on a shared model
Coded to ICD-10, ICD-O-3, and Orpha at capture

Case study, anonymised

One record per child, across every centre.

A national paediatric oncology register runs on the platform: each child's record holds the clinical history with its imaging, pathology, genomic data, and biobank samples, coded to international classifications.

One record

Clinical history
Imaging
Pathology
Genomics
Biobank
Follow-up

Every centre contributes to one record per child
Consent recorded for every registration
Linked with other registers
Incidence, survival, and trend views
Anonymised releases to international research

Read the case study

Where hospitals start.

Start with one collection that matters, then extend the same governed layer across the hospital.

Example
Rare disease registry
Orpha-coded records, consent, and follow-up, shared with European registries on your terms.
See the example
Example
Biobank
Every sample linked to the patient's record, consent, and chain of custody.
See the example
Example
Digital pathology archive
Whole slide images kept with the diagnosis, molecular results, and outcome.
See the example
Example
Health data holder
Describe your datasets and answer access requests from one governed layer.
See the example