Serena Labs
PlatformSolutionsCasesTechnologyInsightsAboutContactEspañol
Sign in

TECHNOLOGY

The engine behind the intelligence.

A modern data architecture built for healthcare's unique requirements — GDPR-native, EU AI Act-ready, and production-proven at 1.5M+ patients.

DATA ARCHITECTURE

The Lakehouse for healthcare.

We built Serena's data layer on a modern Lakehouse architecture with ACID transactions, time travel, multi-tenant isolation, and built-in governance. It's production-proven at scale — over 1.5 million patient interactions processed across 15+ enterprise clients.

Lakehouse Foundation

Not just a data lake — ACID transactions, time travel, and schema evolution for reliable healthcare data.

Multi-Tenant Isolation

Per-client data segregation with encryption at rest and in transit. Aligned with GDPR Article 32 technical and organizational measures.

Unity Governance

Data governance, lineage tracking, and compliance audit trail built in from day one.

Real-Time + Batch

Same data table serves streaming (live updates) and batch (historical analysis) without duplication.

WHY HEALTHCARE NEEDS THIS

Modern architecture for modern challenges.

Regulatory Compliance

  • GDPR Article 30: native processing-activity audit trail
  • GDPR Article 17: right to erasure in seconds (not weeks)
  • GDPR Article 32: encryption at rest + in transit, access controls
  • EU AI Act readiness: documented risk profile per AI agent, human-in-the-loop for consequential decisions

Multiple Disconnected Sources

  • EHRs, imaging, wearables, claims, labs, CRM integrations
  • Schema enforcement with controlled evolution
  • No more data silos or manual reconciliation

Real-Time + Batch

  • Live streaming for alerts and dashboards
  • Historical batch analysis for ML training
  • Same source of truth (no duplication)

Security by Design

  • Encryption at rest + in transit
  • Multi-tenant isolation at the table level
  • Audit trail of every access and transformation

COMPLIANCE

Built for European healthcare.

GDPR

Articles 9 / 17 / 30 / 32. Native audit trail. DSAR-ready within 30 days. EU data residency by default.

EU AI Act

Article 6 awareness. Documented risk per AI agent. Human-in-the-loop for consequential clinical decisions. Transparency obligations met.

ENS

Esquema Nacional de Seguridad — pathway to Spanish public-sector certification (High category) when scope requires.

Data Residency

EU region by default (Frankfurt or Dublin). Multi-tenant segregation. Standard Contractual Clauses for cross-border data flows.

Not a medical device. Serena Labs augments operations and engagement — it does not replace clinical decisions. Compliance posture is documented; full details available under NDA.

PROPRIETARY METHODOLOGY

Why we ship in 45 days.

Most healthcare data platforms take 6-12 months to deploy because every implementation rebuilds the data pipeline from scratch. We ship in 45 days because the pipelines, the integrations, and the agent architecture are already built — refined over 5+ years and 15+ enterprise clients. Your implementation customizes the edges. The engine is ready.

FREE TOOLS

Try the simulators built for your role.

Four interactive business-case calculators for healthcare executives — calibrated against Spain peer-reviewed sources and EU industry research. Run your scenario, save it, share it with a colleague. Free account.

Insurance · Channel & RetentionHospital · Episode & LeakageDigital Health · Unit EconomicsPharma · PSP ROI
Open Free ToolsAlready a member? Sign in

Want to go deeper?

We'd be happy to walk through our architecture under NDA, including data residency options, EU AI Act readiness, and integration patterns for your stack.

Book a Technical Walkthrough

TECHNICAL QUESTIONS

What CIOs and CTOs ask most often

Where exactly is the data stored?
Customer choice between AWS Frankfurt, Madrid, or Ireland. Storage, compute, and embedding indexes all stay in the same region. No data egress to non-EU regions, including for telemetry.
How is multi-tenant isolation enforced?
Row-level security with tenant_id propagated through every query, plus per-tenant encryption keys for at-rest data. Cross-tenant reads are blocked at the database layer, not just the application layer.
What models do you use, and can we bring our own?
We use a mix of frontier and specialized models depending on the task. For sensitive workloads we run open-weight models in-region. BYO LLM is supported via a routing layer when contracts require it.
Is this EU AI Act ready?
Yes. Our agents are designed for the EU AI Act risk-based framework. We maintain Article 16 documentation (data governance, traceability, human oversight) and offer GPAI compliance reporting where applicable.
Can you integrate with our existing CRM, EHR, and data warehouse?
Yes — we ship native connectors for Salesforce, HubSpot, common HL7/FHIR EHR stacks, and modern lakehouses (Snowflake, Databricks, BigQuery). Custom integrations follow a documented patterns library.