Sovereign Infrastructure for Health AI Agents
Built for European regulation, not retrofitted
Loretta's PIaaS platform doesn't just run analytics. It hosts autonomous health AI agents that operate inside your institution, under your governance, with full EU AI Act auditability.
Each agent is a specialist: one orchestrates prevention programmes, another optimises workforce allocation, another monitors polypharmacy risks, another audits all the others for fairness. They share a common property: they execute inside your Trust Centre. Your data stays with you. The agent comes to the data.
These are the agents we are building today. But the infrastructure is open: your institution can deploy additional agents for your own use cases, from clinical decision support to operational optimisation, on the same sovereign, federated platform.
The three verticals below are our first go-to-market applications. The agentic layer is the long-term platform.
Identifies at-risk cohorts, recommends interventions, tracks programme effectiveness.
Predicts staffing demand, monitors burnout risk, optimises training allocation.
Detects multi-drug interaction risks across federated prescription data.
Audits all other agents for equity across age, gender, SES, migration background.
+ Any future sovereign health agent deployed by institutional partners or third-party developers.
The employee sees an app. The manager sees a dashboard. What we deliver is the federated prevention intelligence layer underneath both.
Data stays on-device; employer never sees individual data.
Manager sees cohort-level only; minimum group sizes enforced.
Turn mandated prevention spend into measurable health outcomes.
Loretta enables GKV funds to turn claims data into actionable prevention programmes: identifying which insured populations need intervention, which interventions will work, and how to allocate mandated prevention budgets (§20 SGB V) effectively. All computation happens inside the insurer's own data environment.
The Pflege sector needs infrastructure, not another dashboard.
Patient and workforce data stay inside each facility. Only encrypted model updates are shared.
Loretta's infrastructure is not limited to three verticals. Prevention intelligence applies wherever chronic disease can be anticipated and averted:
Each of these is a future application that runs on the same sovereign infrastructure. The three verticals we are commercialising today were chosen because they have the clearest path to revenue. But the platform serves the full prevention landscape.
In the EU, 85% of deaths are caused by chronic diseases. Yet only 3% of national health budgets go to prevention. The gap between what institutions are mandated to spend and the infrastructure they need to spend it effectively: that is where PIaaS sits.
High-risk classification. Full algorithmic transparency and audit trails (Article 14).
Federated architecture as implementing technology for secondary use provisions.
Germany's health data use law. Secondary use under strict institutional governance.
Medical Device Regulation compliance pathway for clinical decision support.
Privacy by design. No raw patient data ever leaves institutional custody.