VECTORIO“EHDS is on the calendar; your data pipeline is still a folder of scripts”
Hospital data, analysed in-house.
Documented pseudonymisation with verifiable provenance
- EHDS secondary use is coming and our data is not ready for it.
- Patient data cannot go to a US cloud, and our own cluster is small.
- The data protection officer asks how this extract was produced; I have a script.
- Documented pseudonymisation with verifiable provenance
- Clinical-note summarisation inside the hospital
- Works on hospital hardware without a cloud or large cluster
From today to the first result.
an ETL script per project
The data protection officer asks how an extract was produced and the answer is a script, while every project rebuilds the same HL7 and FHIR ETL.
PII is redacted by a documented node, the extract ships with a signed evidence pack, and clinical-note summarisation runs inside the hospital with zero network egress.
Run one extract through the redaction workflow on a copy of the data.
one extractPersonal data masked, with proof. Step by step.
Records are fetched page by page, sealed before masking, masked, and sealed again — two evidence packages that show exactly what changed.
- Page Numbers (2-3) · Build API URL · JSONPlaceholder Posts · Pre-Mask EvidenceSequence Generator · Compute Fields · REST API Subscriber · Evidence Collector
- PII Masking · Pre-Mask Evidence PackData Masking · Evidence Pack
- Post-Mask Evidence · Pre-Mask EPK VerifierEvidence Collector · Evidence Verifier
- Masked Posts CSV · Post-Mask EvidenceCSV Publisher · Evidence Pack
- Post-Mask EPK VerifierEvidence Verifier
page-gen → Generators.Sequence
url-builder → ETL.Compute
api-reader → RestApi.Subscriber
evidence-pre → Forensic.EvidenceCollector
pii-mask → Security.Mask
evidence-post → Forensic.EvidenceCollector
csv-output → CSV.Publisher
epk-pre → Forensic.EvidencePack
verifier-pre → Forensic.EvidenceVerifier
epk-post → Forensic.EvidencePack
verifier-post → Forensic.EvidenceVerifier- Page Numbers (2-3)Counts the pages to fetch.
- Build API URLBuilds each page address.
- JSONPlaceholder PostsCalls the REST API and yields the page rows.
- Pre-Mask EvidenceHashes the raw rows (SHA-256, batches signed every 50) before masking.
- PII MaskingPartially masks titles and hashes bodies.
- Pre-Mask Evidence PackPacks the pre-mask manifest into a sealed evidence package.
- Pre-Mask EPK VerifierRe-computes and checks the pre-mask package.
- Masked Posts CSVAppends the masked rows to the result file.
A real workflow from our end-to-end test suite; the designer view is illustrated.
What you get. From the first pilot.
Documented pseudonymisation with verifiable provenance
Clinical-note summarisation inside the hospital
Works on hospital hardware without a cloud or large cluster
Why your tools fall short.
Cloud health platforms conflict with GDPR/sovereignty expectations
Traditional ETL suites are heavy, costly and weak on audit
Research scripts lack governance
- EHDS timeline (2029/2031)
- GDPR and national health-data laws
- Research data centres (e.g. university-hospital integration centres)
- Local-LLM interest for clinical text without cloud
Your pilot. Three steps.
Your data shape
Tell us what data you have, where it lives and what the result must be.
A scoped pilot
On your data and your hardware, with the success measure written down first.
A clear decision
You see the result against that measure — and keep every workflow and record.
Related roles in Regulated evidence.
Bring us your hardest data problem. Let's solve it together.
A short pilot on your data and your hardware. We agree on what success looks like before we start.