Industry · Healthcare & digital health

Digital health applications and data, secured by confidential computing

Patient data stays encrypted even during processing, keys remain under your control, and every runtime proves its integrity cryptographically — for the hospitals and health insurers (statutory and private) who operate health data, the public-health bodies they serve, and the digital health vendors who build on it.

Talk to an expert →Try the platform →

Statewide medical data platform in pilot (MEDI:CUS)

Privacy-preserving health research (AnoMed)

BUILT FOR HEALTH Insurers, HOSPITALS AND DIGITAL HEALTH VENDORS

Architecture matching gematik's VAU/TEE requirement class

ONE ROOT CAUSE

Care is digitizing, research and AI need the data, and the rulebook is layered — GDPR Art. 9, gematik, NIS2. All three pressures trace to the same root:

conventional infrastructure exposes data during processing.

Confidential computing removes that exposure instead of managing it — encrypted in use, keys outside the operator's reach, integrity attested.

In practice: the KIS periphery and patient services digitized on cloud timelines, the DPO answered with evidence, and research data usable instead of locked away.

Outcomes leadership can commit to.

TWO KINDS OF HEALTHCARE READER

RUN · hospitals, insurances, public health

Operate health data with the operator excluded

Platforms, records and analytics on cloud economics — patient data sealed from every infrastructure operator, attestation instead of assurances.

BUILD · DIGITAL HEALTH VENDORS

Ship products on gematik-grade infrastructure

The trusted-runtime layer gematik's specifications describe — for ePA-adjacent services, telemedicine and TI applications, as buildable infrastructure your architecture review can approve. ePA, eRezept and the sectoral Identity Provider already run on confidential computing — the VAU in gematik's terms.

WHAT HEALTHCARE ORGANIZATIONS BUILD ON IT

Telematikinfrastruktur applications — the pattern gematik already mandates

ePA, eRezept and the sectoral Identity Provider already run on confidential computing — the VAU in gematik's specifications. ePA-adjacent services, telemedicine and TI applications build on the same trusted execution layer, as infrastructure your architecture review can approve.

Health data platforms — hospital to statewide

KIS-adjacent systems, EMR and registry backends on transparently encrypting databases — contents encrypted during query processing, no rewrites. The MEDI:CUS pattern.

Research without raw-data exposure

Secondary use under the GDNG, made technically safe: researchers compute on data they never see in the raw — the AnoMed pattern.

Clinical AI & Confidential AI

Sealed end to end; for external models, a pseudonymization layer tokenizes identities before anything leaves the enclave.

Cross-institution collaboration

Consortia derive shared insight without pooling patient data — custody stays put, verified results move.

Payers: claims & care analytics

Krankenkassen and insurers run claims processing, care-management and ePA-adjacent analytics on member data — sealed, evidenced, and defensible to the DPO and the supervisory authority alike.

Compliance, mapped

GDPR Art. 9, NIS2 for the health sector, gematik where it applies — requirement by requirement, on the dedicated pages.

BUILDING FOR HEALTHCARE?

Your product doesn't change.
Your hardest sales objection does

PLANE 1

Your clinical applications

EMR, telemedicine, imaging, health AI — deployed exactly as they are. No rewrites, no SDK.

PLANE 2

Confidential runtime

Confidential VMs, Kubernetes and encrypting databases — patient data encrypted in use, per tenant, on whatever cloud your customer requires. [PLACEHOLDER — confirm encrypted-database catalog before naming services publicly.]

PLANE 3

Custody & attestation

Your customer holds their own keys; only verified, untampered workloads can decrypt — everything else fails attestation and gets nothing.

code changes

Minutes

to deploy confidentially

console across every cloud

The vendor multiplier

Migrate once, and every tenant's privacy posture upgrades simultaneously

— every deal, tender and DPO review inherits the same answer. A managed operator runs the infrastructure

without ever being able to read the data.

Selling into the telematics infrastructure?

FAQ

What healthcare teams ask

Q·01

Do we need to rewrite our systems or products?

No — workloads and product images run in confidential environments without code changes. Start with the data platform or a new service, not the KIS core.

−

Q·02

How does this change the GDPR Art. 9 position?

Encryption that persists during processing is materially stronger than the at-rest/in-transit baseline — DPIAs get an enforcement claim instead of a promise. The legal assessment stays yours; the premise improves.

Q·03

Are you gematik-approved?

Approval attaches to products, not infrastructure vendors. enclaive provides the runtime layer matching the requirement class gematik describes; your product pursues approval on top.

Q·04

Can clinicians use Confidential AI on patient data?

Sealed inference removes the exposure for self-hosted models; the pseudonymization layer covers external ones. What remains is your governance decision — on a defensible technical premise.

Q·05

We're a vendor — what does this change in our customers' DPIAs and DPA negotiations?

The technical premise: Art. 9 data processed with encryption in use and customer-held keys is a materially stronger position than contractual assurances over readable infrastructure. Your customer's DPO gets an enforcement claim to assess instead of a promise to trust.

Q·06