Prototype · Governance Infrastructure

Structured Comparability
for Cross-Border Healthcare

Mobility exists. Comparability does not. Buiten.ai introduces a supervised governance infrastructure that makes cross-border care structurally intelligible — without standardization, automation, or public ranking.

Buiten.aiGovernance Infrastructure Layer
buitenarts
GP Referral
Gatekeeper continuity
buitenscore
Hospital
Compliance dashboard
buitenclaims
Insurer
Claims governance
ALIGNMENT
VARIANCE
DIVERGENCE
§ Institutional Definition

What Buiten.ai
is — and is not.

This infrastructure is defined as much by its boundaries as by its purpose. Clarity is governance.

Three Pillars of
Structured Comparability

Each pillar addresses a distinct systemic failure in cross-border healthcare — together forming an integrated governance architecture.

01 —

Clinical Comparability

The Indication Matrix structures clinical equivalence across divergent national thresholds — without imposing protocols or overriding physician judgment. Divergence is measured, not judged.

Explore the Matrix →
02 —

Financial Predictability

The Insurance Sustainability Framework transforms volatile cross-border claims into structured, risk-adjusted signals — reducing reserve uncertainty, shortening review cycles, and stabilizing actuarial forecasting.

Insurance Interface →
03 —

Institutional Transparency

Participating hospitals gain structured internal visibility through the compliance dashboard — enabling learning, not exposure. Complexity is contextualized, never penalized. Autonomy remains intact.

Hospital Dashboard →

Role-Separated
Access Portals

Three operationally distinct interfaces — each purpose-bound, role-restricted, and governed by the shared infrastructure architecture.

Governance Signals

AI produces structured, explainable signals. No automated decisions. Every output includes trigger parameter, reference threshold, divergence metric, and risk-adjustment modifier.

Signal Green

Alignment

Clinical escalation falls within structured reference thresholds. Documentation complete. Conservative pathway confirmed.

Signal Yellow

Contextual Variance

Divergence detected. Contextual review recommended. Risk-adjustment modifiers applied before signal generation.

Signal Red

Significant Divergence

Structural divergence exceeds reference threshold. Prioritized human review required. No automated consequence.

The Indication
Matrix

The clinical backbone of the infrastructure. Each procedural cluster is decomposed into measurable decision nodes — compared across systems to quantify structural divergence without declaring error.

The matrix evolves through continuous academic recalibration. It is a living governance instrument, not a fixed protocol.

TDI — Threshold Divergence Index
Proportional deviation from structured reference threshold
High divergence triggers transparency and review recommendation — not automatic sanction. Ambiguity increases oversight intensity, never presumption of misconduct.
// Example: Lumbar Disc Herniation — NL–TR Corridor
Decision NodeNL ReferenceTR KayseriMapping
Conservative therapy duration≥ 6 weeks documented≥ 6 weeks documentedHigh
Neurological deficit documentationMandatory pre-escalationMandatory pre-escalationHigh
MRI confirmation requirementRequired within 3 monthsRequired, timing flexibleModerate
Red-flag symptom criteriaStandardized NICE-alignedPartially standardizedModerate
Escalation timing thresholdDefined in DBC pathwaySpecialist discretion-basedLow — Manual Review
Failed conservative documentationStructured GP recordVariable documentation formatLow — Manual Review
Age-Weighted Baseline
Physiological reserve modeled by age cohort before outcome comparison.
Comorbidity Indexing
Concurrent condition burden integrated into expected outcome baseline.
Frailty Calibration
Frailty markers adjusted to prevent penalizing high-dependency populations.
Case-Mix Normalization
Tertiary centre complexity contextualized — never compared as equivalent to routine care.

Accountability
Before Automation

Designed for high-sensitivity healthcare environments. AI functions as a supervised analytical layer — never as a decision authority.

01 —

Human-in-the-Loop Architecture

Every governance signal is supervised. No reimbursement decision, institutional evaluation, or compliance signal is executed without human oversight. Final decisions remain with insurance institutions, clinical review boards, and governance bodies.

Mandatory SupervisionNo Auto-ApprovalInstitutional Authority
02 —

Explainable AI (XAI) Protocols

No opaque probability scores are delivered without traceable reasoning. Each signal includes its trigger parameter, reference threshold, divergence metric, risk-adjustment modifier, and mapping confidence level. Transparency is structural, not optional.

Full Signal TraceabilityAudit PathwaysVersion Control
03 —

EU AI Act & GDPR Alignment

Designed under high-accountability AI governance principles with GDPR data minimisation standards and purpose limitation requirements. Role-based access control ensures data remains purpose-bound across all operational layers.

EU AI Act PrinciplesGDPR CompliantData Minimisation
04 —

Bias Monitoring & Fairness Controls

Periodic evaluation of signal distribution asymmetry, specialty-level variance, and institutional clustering effects. High-complexity centres treating difficult populations are protected by architectural design — complexity is contextualized, not penalized.

Signal Distribution AuditSpecialty Variance ReviewComplexity Protection
05 —

Academic Oversight Consortium

The Indication Matrix is developed and recalibrated by an academic consortium under periodic review to ensure clinical validity, evidence alignment, threshold updates, and risk adjustment recalibration. Institutions may initiate recalibration dialogue.

Living InstrumentEvidence IntegrationChallenge Rights
06 —

Coordination Without Concentration

Identity data never crosses borders. Encrypted tokenization ensures no personal identifiers are transmitted. Each operational layer is separated by role with restricted data access privileges. Cross-domain aggregation is prevented by architectural design.

Zero Cross-Border IdentityTokenized ArchitectureRole Separation

Kayseri — Dordrecht
Governance Sandbox

The first validation corridor structures an already-existing migration-linked care flow. The pilot does not create mobility — it structures a pattern that exists.

// Active Validation Corridor — Phase 1
Kayseri
Türkiye
~2,700 KM
Dordrecht
Netherlands
01

Patient contacts Dutch family physician

GP evaluates necessity of examination in Kayseri under existing gatekeeper logic.

02

Encrypted referral token generated

No personal identifiers cross borders. Identity and clinical data are architecturally separated.

03

Clinical assessment in Kayseri

Participating clinical partner operates under pre-alignment training and documentation standards.

04

Report returns to originating GP

Clinical report transmitted back to the Dutch GP system. Gatekeeper continuity preserved.

05

Governance signal generated

Indication alignment, complication context, and mapping confidence evaluated under supervised AI architecture.

Phase 1
Active corridor validation in Kayseri–Dordrecht corridor
4
Expansion phases toward multi-corridor European network
0
Personal identifiers transmitted cross-border
Institutional autonomy preserved throughout participation

The corridor measures signal consistency, inter-review agreement, variance reduction, dispute cycle length, and institutional feedback. Expansion follows measurable validation — evidence before scale.

ACTIVEPhase 1 — NL–TR corridor validation
NEXTPhase 2 — Matrix recalibration & institutional onboarding
PLANNEDPhase 3 — Expanded corridor replication within Europe
PLANNEDPhase 4 — Multi-corridor structured governance network

Safeguards
Before Scale

Trust is engineered: supervised signals, privacy-by-design, explainability, audit readiness, and strict role separation across the ecosystem.

01 —

No Automated Decisions

Buiten.ai produces structured governance signals only. Reimbursement and institutional determinations remain under human institutional authority.

Human-in-the-LoopSignal-Based
02 —

Privacy-by-Design (GDPR)

Identity data stays within originating systems. Tokenised workflows minimise data exposure and enforce purpose limitation across borders.

Data MinimisationNo Central Identity
03 —

Explainability & Traceable Signals

Every signal is linked to clear trigger variables (matrix thresholds, divergence logic, risk context) so reviewers can understand what changed and why.

XAITransparent Logic
04 —

Fairness Controls

Risk-adjusted contextualisation (age, comorbidity, frailty, case-mix) protects high-complexity centres from structural penalisation.

Risk AdjustmentComplexity Protection
05 —

Auditability & Version Control

Governance logic is version-controlled and auditable. Changes are documented, reviewable, and traceable—preventing silent updates and undocumented drift.

Audit TrailsChange Logs
06 —

Role Separation & Data Boundaries

Operational layers are separated by design: referrals (buitenarts), institutional dashboards (buitenscore), and insurer review (buitenclaims) each operate with role-based access.

RBACSeparation of Concerns

Who We Are
and How We Operate

Buiten.ai is a governance infrastructure prototype focused on structured comparability — not care delivery, not insurance, not regulation.

01 —

Institutional Definition

Buiten.ai is not a healthcare provider, insurer, or ranking authority. It is a supervised governance layer designed to reduce cross-border ambiguity.

Governance LayerNo Ranking
02 —

Academic Oversight Model

The Indication Matrix is intended to evolve through academic recalibration: evidence integration, threshold updates, and structured feedback pathways.

Living InstrumentEvidence-Led
03 —

Corridor-Based Validation

The infrastructure is validated through controlled corridors. Pilot before scale. Evidence before expansion. No centralisation-by-default.

Pilot FirstScalable
04 —

Institutional Partnerships

Participation is improvement-oriented. Institutions retain autonomy, gain transparency, and access recalibration channels under governance oversight.

Autonomy PreservedTransparency